WO2019175686A1 - Système d'intelligence artificielle et d'administration de routes à la demande - Google Patents

Système d'intelligence artificielle et d'administration de routes à la demande Download PDF

Info

Publication number
WO2019175686A1
WO2019175686A1 PCT/IB2019/051294 IB2019051294W WO2019175686A1 WO 2019175686 A1 WO2019175686 A1 WO 2019175686A1 IB 2019051294 W IB2019051294 W IB 2019051294W WO 2019175686 A1 WO2019175686 A1 WO 2019175686A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
images
detection
parking
objects
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/IB2019/051294
Other languages
English (en)
Inventor
Jayant RATTI
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=67908733&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=WO2019175686(A1) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Priority claimed from US15/919,033 external-priority patent/US10296794B2/en
Application filed by Individual filed Critical Individual
Priority to AU2019235551A priority Critical patent/AU2019235551B2/en
Priority to EP19766866.8A priority patent/EP3676754A4/fr
Priority to SG11201909815R priority patent/SG11201909815RA/en
Priority to CA3093836A priority patent/CA3093836A1/fr
Publication of WO2019175686A1 publication Critical patent/WO2019175686A1/fr
Anticipated expiration legal-status Critical
Priority to AU2021254525A priority patent/AU2021254525A1/en
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
    • G06F18/24133—Distances to prototypes
    • G06F18/24137—Distances to cluster centroïds
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/09—Supervised learning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/94—Hardware or software architectures specially adapted for image or video understanding
    • G06V10/95—Hardware or software architectures specially adapted for image or video understanding structured as a network, e.g. client-server architectures
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
    • G06V20/597—Recognising the driver's state or behaviour, e.g. attention or drowsiness
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/18—Eye characteristics, e.g. of the iris
    • G06V40/19—Sensors therefor
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/20—Movements or behaviour, e.g. gesture recognition
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/02—Knowledge representation; Symbolic representation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/01—Probabilistic graphical models, e.g. probabilistic networks

Definitions

  • the present disclosure relates to crowd-sourced artificial intelligence (AI) and more specifically relates to an artificial intelligence based system and method for determination of traffic violations and objects of visual interest with data obtained from people.
  • AI artificial intelligence
  • cities and towns will often provide metered parking spaces and handicapped parking spaces. Handicapped spaces are designated only for authorized handicapped persons.
  • the parking meters accept coins or dollar bills in return for time on the meter, which represents the amount of authorized time that the vehicle may remain within the parking space. If the vehicle remains within the parking space after the meter expires, the owner of the parked vehicle is subject to citation or towing.
  • the parking garages / parking lots would provide a sheltered or open aired parking space. Sometimes these are multi-level garages, to support multiple cars.
  • the parking lot is claimed by the authorities at a space for the convenience of the public and parking attendants are commissioned / enstated to facilitate parking for shoppers, show owners, visitors etc.
  • the location within the driveway or underground/basement levels and the side-walk location adjoining the establishment are granted to allow for parking.
  • Private parking spaces are also available in areas where parking is in short supply. These parking spaces typically come at some expense to the vehicle owner, but usually offer the convenience of not requiring that the owner pay money throughout the day. Moreover, private parking spaces often offer greater security and a convenient location. Many private garages issue a fixed number of monthly parking passes sometimes costing $200.00-$300.00 or more. Unfortunately, unauthorized persons often park in the assigned spaces resulting in significant inconvenience to both the owner of the private parking lot and the owner of the vehicle which is entitled to the parking space. In order to reduce the unauthorized parking within the assigned spaces, the parking lot owners will often employ expensive attendants to monitor the parking lot.
  • the parking enforcement car drives around the neighborhood and determines if a parking spot is empty or occupied and then the information fed to the central server from the parking pay stations about the expiration of parking times for various spots.
  • Learned Convolutional Neural Networks by Syed Zain Masood, et al, proposes a sighthound’s fully automated license no. plate detection and recognition system.
  • the core technology of the system is built using a sequence of deep Convolutional Neural Networks interlaced with accurate and efficient algorithms.
  • Inc. teaches a parking assist system of a vehicle that includes a camera that, when disposed at the vehicle, has a field of view exterior of the vehicle.
  • An image processor is operable to process image data captured by the camera to detect parking space markers indicative of a parking space and to identify empty or available parking spaces.
  • the system may detect multiple parameters associated with an automobile, and preferences associated with the driver and allocate a parking by generating a parking token corresponding to the parking slot and provide a first set of navigation instruction and orientation instructions for guiding the driver to reach the parking slot based on monitoring of the parking token.
  • jumping red lights or stop signs are considered very severe violations of traffic, posing risk of life and property for everyone within reach of the violation area. Further, other multiple violations are also important to record and cause grave harm to people and property. Examples of some available prior- art literature are as listed below:
  • a paper titled“a system for tracking and annotating illegally parked vehicles presents an automatic method for identifying such events in CCTV video, by first tracking all the related objects and then annotating the events with appropriate keywords for storing and retrieval purposes.
  • the method proposed makes use of a combination of video object tracking algorithms and techniques for capturing knowledge in keyword-based ontology structures. Starting from low level visual information extracted from each video frame, high-level semantics such as moving objects are identified and classified. A probabilistic model, which takes its inputs from the visual modules, is used for identifying illegally parked vehicles. Finally, the keyword ontology, constructed automatically from expert descriptions, is linked to objects identified in the video.
  • a paper titled“Police Eyes” (DOI: l0.l l.09/ECTICon.20l3.6559635, Print ISBN: 978-1-4799-0546-1, Publisher: IEEE, Conference Location: Krabi, Thailand) describes a mobile, real-time traffic surveillance system we have developed to enable automatic detection of traffic violations.
  • police Eyes would be useful to police for enforcing traffic laws and would also increase compliance with traffic laws even in the absence of police.
  • the system detects illegal crossings of solid lines using image processing and efficient computer vision techniques on image sequences acquired from static IP cameras.
  • Chinese Patent application CN101183427A by Chengjun is system that checks the violation of illegal parking by monitoring a parking region continuously from a video stream acquired by an omnidirectional sensor.
  • the system starts an event which comprises of tracking the vehicle using a multi-object tracker, license plate recognition using background subtraction and template matching.
  • This patent doesn’t employ any Deep learning methods.
  • a paper titled“Integrating motion and appearance for overtaking vehicle detection proposes an algorithm for detecting overtaking vehicles using motion cues from the scene. Motion compensation of video data is performed using the optical flow of the scene and epipolar geometry. After post processing and outlier removal, overtaking vehicle candidates are produced.
  • United States Patent US6546119B2,“Automated traffic violation monitoring and reporting system” by Redflex Traffic Systems provides a system for monitoring and reporting incidences of traffic violations at a traffic location is disclosed.
  • the system comprises a digital camera system deployed at a traffic location.
  • the camera system is remotely coupled to a data processing system.
  • the mounted camera system is used for traffic light violation detection.
  • the intersection camera system is strategically placed at an intersection to monitor and record incidences of drivers disobeying a red light.
  • a vehicle is detected approaching the stop line of a monitored lane, it is tracked and its speed is calculated. If the vehicle is detected entering the intersection against the traffic signal, an evidentiary image set is captured.
  • United States Patent US9286524B titled“Multi-task deep convolutional neural networks for efficient and robust traffic lane detection” assigned to University of Technology Sydney, Toyota Motor Corp provides a computing device comprising of one or more processors for controlling operations of the computing device; and a memory storing data and program instructions used by the one or more processors,
  • a paper by Angelova, Anelia, et al. titled“Real-Time Pedestrian Detection with Deep Network Cascades” presents a new real-time approach to object detection that exploits the efficiency of cascade classifiers with the accuracy of deep neural networks.
  • Deep networks have been shown to excel at classification tasks, and their ability to operate on raw pixel input without the need to design special features is very appealing.
  • deep nets are notoriously slow at inference time.
  • US patent application US20170300763A1 titled“Road feature detection using a vehicle camera system” by GM Global Technology Operations LLC provides a computer- implemented method for road feature detection, the method comprising receiving, by a processing device, an image from a camera system associated with a vehicle on a road; generating, by the processing device, a top view of the road based at least in part on the image; detecting, by the processing device, lane boundaries of a lane of the road based at least in part on the top view of the road; and detecting, by the processing device, a road feature within the lane boundaries of the lane of the road using machine learning.
  • a paper by Zeng, Xingyu, et al. titled “Deep learning of scene- specific classifier for pedestrian detection” provides performance of a detector depends much on its training dataset and drops significantly when the detector is applied to a new scene due to the large variations between the source training dataset and the target scene.
  • the authors propose a deep model to automatically learn scene-specific features and visual patterns in static video surveillance without any manual labels from the target scene. They also propose a cluster layer in the deep model that utilizes the scene specific visual patterns for pedestrian detection.
  • a paper by Haloi, Mrinal et al, titled“A robust lane detection and departure warning system.” is based on single camera sensor.
  • For lane detection a modified Inverse Perspective Mapping using only a few extrinsic camera parameters and illuminant Invariant techniques is used.
  • Lane markings are represented using a combination of 2nd and 4th order steerable filters, robust to shadowing. Effect of shadowing and extra sun light is removed using Lab color space, and illuminant invariant representation. Lanes are assumed to be cubic curves and fitted using robust RANSAC. This method can reliably detect lanes of the road and its boundary. This method has been experimented in Indian road conditions under different challenging situations and the result obtained was very good.
  • US patent application US20170372161A1 titled“intelligent automatic license plate recognition for electronic tolling environments” by Accenture implements technical solutions that improve the accuracy of automatic license plate recognition.
  • the IALPR analyzes an image of a vehicle proximate to a toll collection point using optical character recognition (OCR), and determines candidate license plate identifications based, at least in part, on the corresponding OCR confidence level.
  • OCR optical character recognition
  • the IALPR can also perform fingerprinting for candidate license plate images and matching analysis with a knowledge base, resulting in additional confidence levels.
  • the IALPR can also perform behavioral analysis on the candidate license plate identifications, including trip context analysis, historical behavioral analysis, or other analytics.
  • the IALPR can generate an overall confidence level for the candidate license plate identifications responsive to the OCR and vehicle fingerprint confidence levels and the behavioral analysis. This enhanced analysis helps the IALPR reduce the number of incorrect license plate identifications and reduce the need for human review.
  • US patent US8509486B2 titled“Vehicle license plate recognition method and system thereof’ by National Chiao Tung University requires a region where a vehicle license plate image exists is detected according to the edge densities of an input image and a vehicle license plate specification.
  • a text area of the vehicle license plate image is divided into a plurality of character images.
  • the character images are binarized to obtain a plurality of binarized character images.
  • a plurality of characters is recognized from the binarized character images.
  • the characters are recombined to form a character string.
  • the abovementioned steps are repeated to obtain a new character string from another image of the same vehicle, which is captured at a next time point.
  • the character string is compared with the new character string character by character to obtain a comparison result for verifying reliability of recognition through a voting technique.
  • US patent US8184863B2 titled“Video speed detection system” by American Traffic Solutions Inc. provides a system and method for measuring vehicle speeds using video sensing.
  • the system comprises a tracking camera that generates accurately time- stamped video sequences of the monitored road at high frame rates and a processing unit that analyzes the video sequences in real time to detect moving vehicles and calculate their speeds using the time-over-distance method.
  • the system automatically detects moving vehicles in each image frame and derives vehicle positions from a projective mapping established from reference markers on the road. Time information is obtained from the date and time stamp associated with each image frame.
  • the system also allows a user of the system to set a trigger speed, and the system automatically saves a video clip of the vehicle whose speed exceeds the trigger speed, recording the speeding vehicle passing through the monitored traffic zone defined by the reference markers.
  • Chinese patent CN202422420U titled“Illegal parking detection system based on video monitoring by Dalian University for Nationalities” provides a utility model discloses an illegal parking detection system based on video monitoring.
  • the illegal parking detection system comprises an image collector, an image processor, an alarm system and a display, wherein the image collector is used for collecting a video image and outputting a video sequence; the image processor is used for carrying out background modeling on the video sequence collected by the image collector by utilizing a codebook model, obtaining a foreground likelihood information image by adopting a background subtraction method.
  • US patent application US20150235091A1 titled “Lane-line recognition apparatus by Denso Corp” provides an apparatus for recognizing a lane line, comprising of an edge-point extractor configured to extract edge points in an image of surroundings of a subject vehicle including a roadway ahead of the subject vehicle
  • US patent US9305223B1 titled“Vision-based indicator signal detection uses spatiotemporal filtering” by Google LLC provides an autonomous vehicle is configured to detect an active turn signal indicator on another vehicle.
  • An image-capture device of the autonomous vehicle captures an image of a field of view of the autonomous vehicle.
  • the autonomous vehicle captures the image with a short exposure to emphasize objects having brightness above a threshold.
  • EP patent application EP2487454A2 titled“Vehicle length estimation” by Scania CV AB describes a method for determining the configuration of a moving vehicle
  • the invention comprises also a system for determining the configuration of a moving.
  • EP patent application EP2574958A1 titled“Road-terrain detection method and system for driver assistance systems” by Nissan Research Institute Europe GmbH describes a road terrain detection system that comprises a method for classifying selected locations in the environment of a based on sensory input signals such as pixel values of a camera image.
  • the method according to any of the preceding claims, wherein the method applied for a specific road terrain such as "road-like area”, “drivable road”, “ego-lane”, “non-ego-lane” , “non- drivable road”, “sidewalk”, “traffic island”, or “off-limits terrain” is automatically parameterized by using positive and negative samples, which are given by training regions such as polygons.
  • EP patent application EP2578464A1 titled“Video-based warning system for a vehicle” by Honda Research Institute Europe GMBH describes a warning system that can be implemented in any kind of, in order to efficiently detect moving objects.
  • the system utilizes at least one camera for a continuous imaging of the surroundings of the vehicle . Thereby, moving objects can be monitored.
  • a computing unit is programmed to estimate a motion of any moving object based on a pixel motion in the camera image. If a dangerously moving object is detected, a warning unit can be used for issuing a warning signal.
  • EP patent application EP2863338A2 titled“Delayed vehicle identification for privacy enforcement” by Xerox Corp describes a method for recognition of an identifier such as a license plate includes storing first visual signatures, each extracted from a first image of a respective object, such as a vehicle, captured at a first location, and first information associated with the first captured image, such as a time stamp. Then a second visual signature is extracted from a second image at different times and locations
  • BV describes a method in which lane level traffic levels are determined based on traffic camera images.
  • a controller aligns a three-dimensional map with a traffic camera view, and identifies multiple lanes in the traffic camera view based on lane delineations of the three- dimensional map.
  • WO patent application W02015056105A1 titled “Forward-facing multi imaging system for navigating a vehicle” by Mobileye Vision Technologies describes a system and method which use cameras to provide autonomous navigation features.
  • a driver-assist system is provided for a vehicle.
  • WO patent application W02013009697A1 titled “Image -based vehicle detection and distance measuring method and apparatus” by Bendix Commercial vehicle systems describes an image based vehicle detection and measuring apparatus that employs an accelerated, structured, search method is described for quickly finding the extrema of a multivariable function.
  • WO patent application W02014130178A1 titled“A method to detect nearby aggressive drivers and adjust driving modes” by Google Inc. describes a computing device may be configured to receive sensor information indicative of respective characteristics of vehicles on a road of travel of a first vehicle. The computing device may be configured to identify, based on the respective characteristics, a second vehicle that exhibits an aggressive driving behavior manifested as an unsafe or unlawful driving action. Also, based on the respective characteristics, the computing device may be configured to determine a type of the second vehicle. The computing device may be configured to estimate a distance between the first vehicle and the second vehicle.
  • EP patent application EP3038012A1 titled“Signal for identifying traffic lights for computer vision” by Kaparazoom SLU describes a method for identifying traffic lights.
  • the invention relates to a signal for identifying traffic lights for pedestrians, using an electronic urban guidance device with computer vision, for blind or visually impaired people. Said identification signal can be used to identify other types of urban elements.
  • US patent US9779314B 1 titled“Vision-based detection and classification of traffic lights” by Waymo LLC describes an autonomous vehicle having a vehicle control system.
  • the vehicle control system includes an image processing system with the image processing system providing instructions to control the autonomous vehicle based on the particular candidate portion representing an illuminated component of a traffic light.
  • US patent US9428192B2 titled“Vision system for vehicle” by Magna Electronics Inc. describes a vision system for a vehicle includes a forward facing camera configured to be disposed at a windshield of a vehicle so as to have a forward field of view through the windshield of the vehicle.
  • the forward facing camera is operable to capture image data for an adaptive speed control system of the vehicle.
  • US patent US7720580B2 titled“Object detection system for vehicle” by Magna Electronics Inc. describes an imaging system which includes an imaging array sensor and a control.
  • the image array sensor comprises a plurality of photo-sensing pixels and is positioned at the with a field of view exteriorly of the vehicle.
  • US patent application US20120148105A1 titled “Automated license plate recognition system and method using human-in-the-loop based adaptive learning” by Xerox Corp. describes an automated license plate recognition (ALPR) system and method using a human-in-the-loop based adaptive learning approach.
  • ALPR automated license plate recognition
  • One or more images with respect to an automotive vehicle can be segmented in order to determine a license plate of the automotive vehicle within a scene.
  • An optical character recognition (OCR) engine loaded with an OCR algorithm can be further adapted to determine a character sequence of the license plate based on a training data set.
  • a confidence level with respect to the images can be generated in order to route a low confidence image to an operator for obtaining a human interpreted image.
  • the parameters with respect to the OCR algorithm can be adjusted based on the human interpreted image and the actual image of the license plate.
  • a license plate design can be then incorporated into the OCR engine in order to automate the process of recognizing the license plate with respect to the automotive vehicle in a wide range of transportation related applications.
  • US patent US7363133B2 titled“Lane detection system and method” by Valeo GmbH describes a lane detection system for a vehicle, having a camera for sensing the carriageway markings in front of the vehicle, and an evaluation unit for evaluating the data collected by the camera.
  • WO patent application WO2016069253A1 titled“Augmented lane detection using kinematic data” by TRW Automotive U.S. LLC describes a system and method provided for detecting the departure of a vehicle from a set of land boundaries.
  • a boundary determination component is configured to determine an associated set of lane boundaries for a vehicle.
  • WO patent application W02007023103A1 titled“Lane Departure Warning And/or Lane keeping system” by Stephan Voltz describes a lane departure warning and/or lane keeping system for a motor vehicle comprising a sensor unit, which is oriented in the travel direction of the motor vehicle and is used for identifying traffic lane(s).
  • US patent US8587649B2 titled“Lane departure warning system” by Create Electronic Optical Co Ltd describes a lane departure warning system (LDWS) installed on vehicles is revealed.
  • the LDWS includes a camera that captures road images and the data of images is sent to an electronic control unit (ECU) for processing and recognition.
  • the ECU is directly connected with a global positioning system (GPS) that provides vehicle speed signals so as to check whether dangerous driving occurs.
  • GPS global positioning system
  • US patent US9760806B 1 titled“Method and system for vision-centric deep- leaming-based road situation analysis” by TCL Research America Inc. describes a method and system for vision-centric deep-learning-based road situation analysis are provided. The method can include: receiving real-time traffic environment visual input from a camera; determining, using a ROLO engine, at least one initial region of interest from the real-time traffic environment visual input by using a CNN training method
  • KR patent application KR20170105845A titled“Driving assistant apparatus, distance detection method and width of lane detection method” by Hyundai Auto describes a system and method for generating a front image containing the other vehicle which is located at the front by a front camera and; And to a driving support apparatus for a control unit for the actual characters on the basis of distance information controls the speed or traveling direction of the vehicle.
  • EP patent application EP3143607A1 titled“Systems and methods for curb detection and pedestrian hazard assessment” by Mobileye Vision Technologies (Jerusalem) Ltd describes a system for a vehicle to identify at least one edge line candidate as an edge line of the curb.
  • EP3261017A1 titled“Image processing system to detect objects of interest” by Delphi Technologies Inc. describes a method of detecting objects of interest in a vehicle image processing system using convolutional neural networks (CNNs)
  • WO patent application WO2017190574A1 and Chinese Patent CN105975929A, titled“Fast pedestrian detection method based on aggregation channel features” by Peking University Shenzhen graduate School describes a fast pedestrian detection method based on aggregation channel features, which comprises an early-stage position calibration process and a later-stage position screening process.
  • the classifier can automatically select features with a good recognition ability to serve as a pedestrian judgment basis.
  • WO patent application WO2017136578A1 titled“Partially occluded object detection using context and depth ordering” by Nissan Motor Co., Ltd. describes a system and method for verifying detection of partially occluded objects (e.g., pedestrians) in the vicinity of a vehicle.
  • An image input device captures an image and/or video of surroundings.
  • CN patent application CN106845430A titled “Pedestrian detection and tracking method based on accelerated area Convolutional Neural Network” by Donghua University describes a pedestrian recognition and tracking method based on an accelerated area Convolutional Neural Network where a pedestrian in an infrared image can be effectively recognized, and real-time tracking for a pedestrian target in an infrared video can be achieved.
  • CN patent application CN107154045A titled“Computer-vision-based method for localizing vehicle with peccancy pressing-line behavior in traffic flow diversion region” by Original Assignee Nanjing University of Aeronautics and Astronautics describes a method for localizing a vehicle with a peccancy pressing-line behavior in a flow diversion region. With the method disclosed by the invention, detection of line pressing by a vehicle in congestion can be realized
  • CN patent application CN106935035A titled“Illegal parking vehicle real-time detection method based on SSD neural network” by Xi'an University of Electronic Science and Technology describes a vehicle real-time detection method based on an SSD neural network, and mainly solves the problems of low detection accuracy and weak robustness under the condition of complex roads and changeable illumination of weather.
  • CN patent application CN106327878A titled “Movable illegal parking snapshot system and implementation method” by Tianjin Zhonghuan System Engineering Co., Ltd. describes a movable snapshot system and an implementation method when an behavior is established as total time of a vehicle target in an area exceeds a time threshold value preset by a system.
  • CN patent application CN107491753A titled “Parking violation detection method based on background modeling” by University of Electronic Science and Technology describes a detection method based on background modeling, firstly inputted video traffic monitoring dynamic background modeling, monitoring of the scene to obtain the background image;
  • CN patent application CN106652465A titled “Method and system for identifying abnormal driving behavior on road” by Chengdu Tongjia Youbo Technology Co., Ltd. describes a method and system for identifying an abnormal driving behavior on road through vehicle identification and a driving direction expressed by an extracted lane line to determine a vehicle driving against the traffic.
  • India Patent Publication 52/2017 titled“Automatic Helmet and Number Plate Detection System” provides automatic helmet and number plate detection system.
  • This invention detects helmet and number plate on the vehicle along with the correct number plate format by taking account the vehicle type i.e. LMV, HMV or trucks.
  • the invention uses Mega-pixel cameras placed at a distance of 50-100 m to capture the high-resolution images and extracts the number plate characters from the number plate.
  • Chinese Patent CN106372662 by Tencent Technology titled“Helmet wearing detection method and device, camera, and server”, comprises: obtaining a scene video image, detecting the scene video image to obtain human position through a human position detection model built by training learning, and determining whether the human position is located in a helmet wearing area or not;
  • a paper titled“classification of motorcyclists not wear helmet on digital image with backpropagation neural network”, by Sutikno et al provides a system that was built is divided into two parts, namely training process and testing process.
  • Backpropagation neural network architecture of this system consists of 400 inputs, one hidden layer consists of 40 neurons, and one output.
  • a paper titled “Traffic Sign Recognition with Multi-Scale Convolutional Networks” by Sermanet et. al provides in traditional ConvNets, the output of the last stage is fed to a classifier. In the present work the outputs of all the stages are fed to the classifier. This allows the classifier to use, not just high-level features, which tend to be global, invariant, but with little precise details, but also pooled low level features, which tend to be more local, less invariant, and more accurately encode local motifs.
  • CNN Convolutional Neural Networks
  • Convolutional Networks by Mrinal Haloi, provides a deep network consists of spatial transformer layers and a modified version of inception module specifically designed for capturing local and global features together. Spatial transformer network capable of generating automatic transformation of input image is used to make classification more robust and accurate along with a modified version of GoogLeNet.
  • United States Patent Application US20130049988A1 by Marc Roeber et al. titled“Device and method for traffic sign recognition” provides a method and apparatus for determining whether to alert a driver of a vehicle to the presence of a traffic sign in the vehicle's driving environment.
  • European Patent EP1096457B1 by Volkswagen Aktiengesellschaft titled “Method and device for electronic recognition of traffic road signs”, provides a method and a device for electronic recognition of road signs are in which the traffic signs detected with a arranged in the motor vehicle electronic camera and interpreted electronically via pattern matching method and displayed in the vehicle and / or act on the automatic speed control of the motor vehicle.
  • United States Patent US6560529B1 by Holger Janssen titled“Method and device for traffic sign recognition and navigation”, provides a method and a coupled system for road sign recognition and for navigation is proposed, which enables a bidirectional data transmission between the road sign recognition device and the navigation device.
  • United States Patent Publication US8983136B2 by Ricoh Company, Ltd. titled “Traffic sign detecting method and traffic sign detecting device”, provides a method and a device for detecting traffic signs in an input image camera.
  • the method comprises a color space converting step of converting the input image into a HSV color space image; a filtering step of filtering, based on a predetermined pass range of a standard color of each of the traffic signs, the HSV color space image to obtain a filtered image, and then generating one or more connected domains based on one or more regions in the filtered image;
  • United States Patent Publication US20110109476A1 by Mitsubishi Electric Research Laboratories Inc. titled“Method for Recognizing Traffic Sign”, provides a method recognizes a set of traffic signs in a sequence of images acquired of a vehicle environment by a camera mounted in a moving vehicle by detecting in each image, a region of interest (ROI) using a parameter space transform “same” or“different.” This enables construction of an efficient multi-class classifier.
  • ROI region of interest
  • Learned Convolutional Neural Networks by Computer Vision Lab, Sighthound Inc., Winter Park, FL, provides an end-to-end license plate detection and recognition method using novel deep CNNs. They claim it to be not only computationally inexpensive, but also it outperforms competitive methods on several benchmarks after conducting successful experiments on leading benchmarks.
  • Chinese Patent CN106250892 by Wan Yongxiu titled“Automatic recognition method for license plate characters”, provides the following steps of inputting a colorful vehicle image; obtaining current light intensity; if the current light intensity is higher than preset light intensity, adopting a license plate location algorithm based on color point pair search and mathematical morphology to carry out license plate location on the colorful vehicle image to obtain a license plate image.
  • Chinese Patent Publication CN106845480 titled “Method for recognizing license plate from picture”, provides a method, the license plate positioning is realized through combination of color positioning and Sobel positioning; through carrying out color judgment on the picture of the license plate, carrying out binaryzation and extracting a character outline and a circumscribed rectangle, the character segmentation is realized; and the characters of the license plate are recognized through a neural network algorithm.
  • United States Patent Publication 20130311075A1 filed by Continental Automotive Systems, Inc. teaches a safety system for motorcycle comp-rises at least one sensor mounted to the motorcycle to sense a feature of an environment surrounding the vehicle.
  • An electronic control unit is configured to receive a signal from the at least one sensor and determine a probability that a safety incident may occur based upon the at least one feature.
  • United States Patent US9704060B2 by Feiyue Wang titled “Method for detecting traffic violation”, relates to a technical field of traffic monitoring, and more particularly to a method for detecting traffic violation.
  • the present invention includes firstly localizing vehicle salient parts through salient features including vehicle license numbers and vehicle rear lights, and representing a vehicle with the vehicle salient parts, then tracking the vehicle with a Kalman filter based on the vehicle salient parts, and finally detecting vehicle violation through moving trajectory analysis and setting violating detecting areas.
  • US patent application 15/689350 titled“On-demand Roadway Stewardship System”, filed by Jayant Ratti, provides an on-demand roadway stewardship system with video reporting features is disclosed.
  • the invention described herein is comprised of a system that allows users with mobile device cameras to record and report roadway safety incidents, traffic violations, crimes and infrastructure problem. Users are encouraged to become stewards by engaging in the system’s rewards program.
  • An on-demand style cloud infrastructure is presented which speeds up video processing and citations. Objects of the invention are to enhance safety and increase public participation in safety.
  • Another object of the present innovation is the method of placing the AI based system as a module to the Stewardship network to act as a specialist or an auditor or a data collector.
  • Figure 18 shows the AI unit whereas other spots are being used by humans to co work along with the AI. As AI continues to become mature, the need for humans in the network would reduce.
  • An example formation of the BPO style model for the traffic violation processing could have several specialists who identify and categorize the traffic violations as seen on the footage captured for cameras on the field.
  • a specialist is by role, either a computer software, or machine or a person who reviews the incoming videos for potential traffic law violators. These specialists can cover one or more of various categories of violations such as: wrong driving, wrong parking, commercial violation, vehicle fitness, among others; as previously mentioned the specialists can be a hybrid mix of humans and Artificially Intelligent Software/Machines.
  • the specialists categorize the videos and issue citations the citations/violations are reviewed by auditors to ensure correctness and fairness.
  • the auditors may themselves be a hybrid mix of software/machines and persons.
  • Another objective of the innovation is to classify various visual features of interest on the road.
  • the AI based system would have a detection module that recognizes a region of interest in an image and classifies it as a vehicle or a person or static/dynamic markings or a policeman.
  • Another objective of the invention is the identification of the region of interest
  • ROI of a vehicle in a series of images along with at least one among the ROIs of road signs, road markings, policeman, person, in the same series of images is sufficient to determine a violation of motor vehicle laws/guidelines.
  • Another objective of the disclosure is the recognition of the ROI of a vehicle along with at least one among the ROIs of road signs, road markings, policeman, person, in the same image is sufficient to determine a violation of motor vehicle laws/guidelines.
  • Another objective of the disclosure is the disclosure of the AI based system, wherein the region of interest (ROI) of a vehicle in a single image is sufficient to determine a violation of motor vehicle laws/guidelines.
  • ROI region of interest
  • Another objective of the disclosure is of an annotation module configured to read license plates captured by the detector module.
  • Another objective of the AI based system to include an occlusion module, which hides objects ROI or hides non-ROI information in the image or series of images or video. This is done in some cases for privacy of certain individuals, in other cases to bring to attention only the concerned areas which are of interest. In some cases this is helpful in making aware publically the kinds of violations and perpetrators. Such can be used in driving training schools and for the courts when the citation is challenged by the violator
  • the detection module which is further configured to detect traffic violations based on the orientation and position of the vehicle. For example the vehicle may be driving on the wrong side of the road; the vehicle may be parked in an irregular manner not conforming to the orientation of the said parking area or block traffic thereby, among other reasons.
  • the detection module further takes into account the position, orientation and location of other vehicles, relative to the vehicle of interest to determine traffic violations. For instance if a vehicle is trying to tailgate another vehicle, or is trying to overtake another vehicle very closely or in a dangerous manner or is trying to squeeze another vehicle aggressively to force the other to brake, etc. are forms of aggressive driving which need to be monitoring using the AI system.
  • Another objective of the AI based system is to add a calculation module configured to calculate a speed of said object utilizing object detection from the location of the said object in different frames of the video or series of images and the relative velocity of the vehicle installed with the said data collector.
  • the vehicle could be moving or stationary, its velocity would be computed based on information received from the GPS and other inertial measurement sensors on the mobile device.
  • Another objective of the system further comprises: an allocation module configured to allocate, one or more parking slot options, based on detection of said parking slots or by matching the host of the parking spot to the driver in need of a parking spot.
  • the owners of the parking spaces can list their availabilities and the illegally parked car owners can be notified regarding their violation and then matched to a parking spot available nearby.
  • Another objective of the system is to connect/partners with taxi service provider companies to provide limited access to the data of drivers or in another aspect, every challan linked to the vehicles will be served to the company owning such vehicles.
  • Another Objective of the disclosure is the ability for vehicle drivers / motorists to lift and drop-off their friends, or friends of friends and other acquaintances from a point A to point B without the need for a commercial drivers’ license or commercial permit.
  • the aiding of the people can be termed as Ubering (coined form the business model of the company UberTM .
  • UberTM coined form the business model of the company UberTM .
  • any driver transporting other people for money is considered a commercial transaction and hence requires a commercial driver’s permit.
  • giving ride to acquaintances for a charge can circumvent a purely commercial activity hence keeping laws of the land in place. Such is already in existence with friends sharing cars and car-pooling to go to similar destinations, or even as a gesture of good faith.
  • Another objective of the disclosure is to devise a method for accumulating training data from specialists by having them annotate license plate numbers of a vehicle, creating bounding boxes around vehicles, identifying the types, make, model and other information about the vehicle and also various violations about the vehicle. The information along with computer logic, would determine the type and severity of a violation.
  • Another objective of the disclosure is for accumulating training data by monitoring GPS and odometric data from the data collectors’ sensors automatically from the sensors available to the data collector on their mobile camera device. This information can be used for various training purposes such as over speeding training among others.
  • Another objective is to create a Stewardship network, where in the system is a game, such that the data collectors, specialists who review the video of series of images collected by the data collectors and auditors who review the work of the specialists, earn points based on the volume and accuracy of the detection of traffic violations. Such points and scores can be used for purchase of goods, merchandize, online shopping, etc. Auditors can also be other specialists in an exemplary example, or vice-versa.
  • Another objective of the disclosure is the ability for users to pay using their points or other digital means to purchase add-ons within the program (such as automated challenging feature, priority assessment of their captured violations, etc.), also for being able to pay for their challans/citations received by the authorities
  • Another objective of the systems is the ability to switch to digital currencies like Bitcoin or such points being awarded in Internet Currencies or Crypto currencies such as Bitcoin, Ethereum, Ripple, among others to be flexible for spending across the world without frequent currency exchanges and other benefits of the cryptocurrencies in general.
  • Another objective of the system is the ability for a simplified way to capture an event through a remote trigger among other gesture triggers possible to trigger the capture of the same.
  • the input module which records the manual gesture of a person such as a button press, or a touch among others with a transmitter module, which transmits the said gesture to the data collector and some mounting hardware which mounts the gesture recognition device to another object or person.
  • the same gesture sensor would be powered by a cell / battery or other power sources.
  • the presence of a remote gesture sensor allows for several benefits including allowing the driver to have their hands on the wheel and not handle the mobile phone; which is illegal / unlawful in many cities.
  • the system includes a non-transitory storage device having embodied therein one or more routines and one or more processors coupled to the non-transitory storage device and operable to execute the one or more routines.
  • the one or more routines include a receiving module that enables to receive at least an object detection signal from one or more vision sensing systems, said object detection signal comprises said images associated with said objects; a determining module, that enables to determine for said objects, a region of interest (ROI) selected from the received images; a detection module that enables to detect a real-time location, a real-time position, a real-time movement, and a category associated with said objects contained in the received images.
  • ROI region of interest
  • a capturing module that enables to capture one or more traffic violations selected from a combination of a motorists not wearing a helmet while driving, a vehicle carrying goods or loads beyond the bounding box of the vehicle, vehicles having no side mirror installed or side mirror not open, more than two people riding on a two wheeler, broken reflector, and non-standard license plates.
  • the data collector can also collect data from the front camera of the mobile phones there by providing information on the facial/body dynamics of the passengers and driver inside a car.
  • the benefit could be optional but greatly add to the safety of drivers and passengers in case of night time driving, or in case of long driving conditions, where drowsiness can cause alarming consequences to life and property.
  • An aspect of the present disclosure relates to a method for detecting objects in images and videos using a deep neural network wherein the method can include the steps of receiving, by a system, at least an object detection signal from one or more vision sensing systems, said object detection signal comprises said images associated with said objects; determining, by the system, a region of interest (ROI) selected from the received images for said objects; detecting, by the system, a real-time location, a real-time position, a real-time movement, and a category associated with said objects contained in the received images.
  • ROI region of interest
  • the method further includes a method for accumulated training data from specialists by having them annotate license plate numbers of a vehicle, creating bounding boxes around vehicles, identifying the types, make, model and other information about the vehicle and also various violations about the vehicle.
  • the proposed system and method has an ability to capture both the road data and the road side data.
  • roadside advertising such as sign boards, traffic signals, notices, and more and directly correlating that to the direction of Gaze of the passengers and driver in the vehicle.
  • the cross-correlation can be used to determine whether the signs on the roads / road-sides are coming into the gaze of the people, alternatively if they are effective in delivering the message to the drivers and passengers.
  • the road side hoardings are advertising the use and sale of Gasoline at a Gas Station, by identifying the direction of gaze from one of the cameras and cross-matching the same with the position of the same advertising in the view field of the phone and camera, it can be determined to a reasonable accuracy / probability that the driver and/or passenger is looking at that particular advertisement.
  • another embodiment of the same feature extends to road safety applications, where road-side traffic warnings or road-safety symbols are present.
  • the direction of gaze, the motion of the vehicle along with the position of the signs in the view field of the camera, when combined can estimate reasonably, whether the driver / passenger is observing the road signs / warnings / traffic symbols. This can lead to improvement in the location and orientation of such signs for maximum viewing efficacy.
  • the present invention proposes utilizing deep convolutional neural networks and machine vision based algorithms like image segmentation to perform a task of detection and recognition to provide complete solution to safe, legal and comfortable parking, driving and riding system for commuters on the roadways.
  • Parking management systems and vacancy indication services in particular, can play a valuable role in reducing traffic and energy waste in large cities. Also proper driving on the roads can lead to faster and safer commute. Since the advent of modem, engineered roadways in the early l800s, local, state and federal authorities have continued working on ways to curb roadway violations.
  • Embodiments of the present disclosure provide a technological solution to a major and common problem of inefficient and illegal parking.
  • the present invention proposes a new technique/mechanism and a complete solution to safe parking. It is a crowd-sourced and crowd-funded on-demand field.
  • the present invention has comprehensively structured our approach based on both manual and automated aspects of the model. Further, the side benefits of the solution are radical. It will drastically improve frequent decongestion of traffic and road management system. The haphazard conditions of roads, which not only cause inconvenience but also are sources of environmental pollution, are improved.
  • the manual aspect of our proposed model provides monetary benefits to the people and could prove to be a source of easy supplementary income.
  • the automated process alone would be beneficial only to organized and standard situations. countries like India, however, have high percentage areas that are unorganized and unsystematic.
  • a real-time Illegal Parking Detection System can be built; it can be a very robust method of single shot multi-box detector algorithm is used for detection module that gives better results than other widely used algorithms like Faster RCNN and YOFO.
  • the input of the system can be a real-time signal or a recorded / captured data coming from the data collectors with a camera. Since the illegal regions in different scenarios varies, the ROI need to be marked manually.
  • the images are provided as input into a trained SSD network to detect the location and category of all vehicles contained in the image by setting a suitable hyper-parameter.
  • the present invention redesigns the aspect ratios of the default boxes in original SSD.
  • the system analyzes the state of the vehicles being tracked during the whole tracking process.
  • the distance that the vehicle moves in the region of interest is obtained by making the difference between the positions of the adjacent two frames. If the difference between the two positions is greater than a given threshold, it is considered to be moving.
  • the alarm will be triggered when the stop time of any vehicle exceeds the given threshold which is being calculated simultaneously.
  • speed calculation of the vehicle detected helps to identify each vehicle video and finds their corresponding distance covered in consecutive frames.
  • the proposed invention tracks each vehicle (or their license plates) and traces their centroid in upcoming frames to get the distance travelled by that vehicle.
  • the proposed invention uses the array of structure to store centroid of each vehicle. In this as vehicle is arrived into the region of interest in the video their corresponding bounding box is created by the detection algorithm, and then the proposed invention generates the centroid of the bounding box.
  • the present invention provides a detection method that utilizes SSD technique to detect vehicles. This includes the use of deep convolutional neural networks and vision-based analysis. Later performing frame by frame evaluation, the present invention calculates speed of the vehicle using pixel level analysis and relating the actual distance to the pixel distance captured by a camera.
  • the first task is to determine if it contains a license plate.
  • the present invention runs the image through our deep license plate detection network. As explained above, this network is robust to different conditions (e.g. lighting changes, viewing angles, occlusion, etc.) and license plate variations (e.g. shape, format, background, etc.), making it highly accurate across a range of scenarios.
  • the present invention needs a mechanism by which the present invention can read the plate.
  • the best approach is to segment/isolate each of the license plate characters, perform character detection and recognition for each individual character before combining the results.
  • the present invention proposes a visual parking vacancy indicator system that utilizes deep convolutional neural networks for detection of vacant and occupied parking slots. The robustness of the detection system is ensured by testing it on a large dataset of labeled parking spots.
  • the detection module is responsible for reporting the occupancy status of a parking stall given the image of the stall.
  • the present disclosure uses a convolutional neural network for this task. Having few constraints on the input data put a heavy burden on the detection system. Therefore robustness and generality of the detection algorithm have the highest priority. This is useful to avoid the space owner to continuously and repeatedly put his empty spot for renting, at sight of the empty spot the space is automatically listed and the space owner get paid the rent for the time their space was used.
  • VGGNet-F has five convolutional layers where each is followed by a pooling layer and Rectified Linear activation function. It has three fully connected layers at the end that use the features from the convolutional layers for classification.
  • the VGGNet architecture features a simple and uniform design throughout the network. Filters’ kernel sizes are all 3 beside first two that are 11 and 5 respectively to reduce the network’s size.
  • VGGNet-F which is the smallest of the VGG networks family.
  • the keeper can provide add-ons to the parker.
  • the present invention can be linked with Google maps and could use interactive interface to guide our parker the shortest way to the parking spot.
  • the parker and the keeper are connected through server using their PARKSAFE account.
  • Parker receives information about the keeper and so the parking spot. Also through the same the keeper gets information about the Parkers in need for a parking space.
  • the proposed invention provides attempts to provide a technological solution to a major and common problem of catching riders riding two wheelers like motor bikes, scooters, scootys etc. (i) who are not wearing a helmet (either driver or passengers) (ii) who are riding with more than 2 people (iii) who are carrying large loads on their scooter or bikes (iv) who do not have a side mirror installed on their vehicles, among others that constitute a violation as well
  • a convolutional neural network is a variant of feed forward neural networks using back propagation algorithm. It learns high-level features from the spatial data like image.
  • the convolutional neural network training consist of convolution layers, ReLu layers Max-Pooling layers, fully connected layers and a loss function (e.g. SVM/Softmax) on the last (fully-connected) layer.
  • a CNN model is built using these images to separate the motorcyclists from other moving objects.
  • the present invention cropped only the top one fourth part of the image as that was the region where the motorcyclist’s head is located most of the time.
  • the present disclosure perform a simple count information which gives the total head count on the two wheeler hence the present invention are able to detect no. of people riding on the two wheeler.
  • the present invention applies a basic threshold to generate a signal/indication when there are more than two riders present on the two wheeler.
  • the present invention builds a CNN model in order to separate the without- helmet from the with-helmet images.
  • This model is trained for the binary classification of helmet and head.
  • the present invention utilizes detections generated in the first aspect of the invention (recited in the detection of motor-cyclists without helmet using CNN). According to this aspect, before distinguishing between helmet and non-helmet wearers, the present invention detects two wheeler vehicles. Therefore after the vehicle detection present invention performs the second aspect of our experiment.
  • the present invention accordingly builds a CNN model to perform binary classification to distinguish two wheelers carrying heavy load from the ones not carrying.
  • the model is trained on a dataset containing two wheelers carrying heavy load (positive sample) as well as the ones not carrying large loads (negative sample).
  • the present invention annotates data and mark bounding boxes over the side view mirror. Once annotated the present invention have complete data set.
  • the present invention again takes off our training from the detections achieved in the first aspect.
  • the detected two wheelers are then checked for presence of side view mirror.
  • common detection algorithms like YOLO/ Faster RCNN the present invention move forward with side view mirror detection.
  • Figure 1 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 2 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 3 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 4 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 5 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 6 illustrates exemplary motor vehicle violations for Category 1 of the violation types, where all information regarding a violation or deviation from the rules is within the picture of the vehicle and rider together only.
  • Figure 7 illustrates exemplary motor vehicle violations for Category 2 of the violation types, where all information regarding a violation or deviation from the rules is captured in a single image showing the vehicle, driver and other roadway information such as signs, markers, objects.
  • Figure 7 illustrates exemplary motor vehicle violations for Category 2 of the violation types, where all information regarding a violation or deviation from the rules is captured in a single image showing the vehicle, driver and other roadway information such as signs, markers, objects.
  • Figure 8 illustrates exemplary motor vehicle violations for Category 2 of the violation types, where all information regarding a violation or deviation from the rules is captured in a single image showing the vehicle, driver and other roadway information such as signs, markers, objects.
  • Figure 9 illustrates exemplary motor vehicle violations for Category 2 of the violation types, where all information regarding a violation or deviation from the rules is captured in a single image showing the vehicle, driver and other roadway information such as signs, markers, objects.
  • Figure 10 illustrates exemplary motor vehicle violations for Category 3 of the violation types, where all information regarding a violation or deviation from the rules is captured in a series of images (or video) showing the vehicle, driver and other roadway information such as signs, markers, objects etc.
  • Figure 11 illustrates exemplary motor vehicle violations for Category 3 of the violation types, where all information regarding a violation or deviation from the rules is captured in a series of images (or video) showing the vehicle, driver and other roadway information such as signs, markers, objects etc.
  • Figure 12 Shows a remote clicker that records the gestures of a person through the push of a button or touch and transmits the trigger to the phone. Shows various locations of how it can be mounted
  • Figure 13 Shows the remote gesture recorder and means of attaching it to different objects or body parts.
  • Figure 14 Shows how the phone can capture the sight direction of people in the car and the location of the advertising boards, traffic signs and other displays on the road. The system determines if the visual markers are indeed watched by people, can rate their quality of placement or determine if they are distracting, causing accidents / bad driving
  • Figure 15 Shows the flow chart of a typical application and shows how a person can use an application to simultaneously act as a specialist and also help collect data for AI Training for automated Specialists.
  • Figure 16. is a Neural Network architecture showing how images are received with ROIs and then the outputs are trained using convolution and sub-sampling. Various trained data is then used for generating the various outputs for automated detection
  • Figure 17. shows how the specialist can do their job on a mobile phone as well
  • Figure 18 Shows other objects of interest being categorized by specialists on their mobile or web devices and such is then sent to the AI server for training automated detection
  • Figure 19 Is taken from the parent patent wherein the on-demand, crowdsourced, stewardship system is shown with specialists, auditors and data collectors connected together through the cloud.
  • Figure 20 Shows the parking allocation scheme for the PARKSAFE technology disclosed in this patent.
  • Embodiments of the present invention include various steps, which will be described below.
  • the steps may be performed by hardware components or may be embodied in machine-executable instructions, which may be used to cause a general-purpose or special- purpose processor programmed with the instructions to perform the steps.
  • steps may be performed by a combination of hardware, software, and firmware and/or by human operators.
  • Embodiments of the present invention may be provided as a computer program product, which may include a machine-readable storage medium tangibly embodying thereon instructions, which may be used to program a computer (or other electronic devices) to perform a process.
  • the machine-readable medium may include, but is not limited to, fixed (hard) drives, magnetic tape, floppy diskettes, optical disks, compact disc read-only memories (CD-ROMs), and magneto-optical disks, semiconductor memories, such as ROMs, PROMs, random access memories (RAMs), programmable read-only memories (PROMs), erasable PROMs (EPROMs), electrically erasable PROMs (EEPROMs), flash memory, magnetic or optical cards, or other type of media/machine-readable medium suitable for storing electronic instructions (e.g., computer programming code, such as software or firmware).
  • Various methods described herein may be practiced by combining one or more machine-readable storage media containing the code according to the present invention with appropriate standard computer hardware to execute the code contained therein.
  • An apparatus for practicing various embodiments of the present invention may involve one or more computers (or one or more processors within a single computer) and storage systems containing or having network access to computer program(s) coded in accordance with various methods described herein, and the method steps of the invention could be accomplished by modules, routines, subroutines, or subparts of a computer program product.
  • Figure 1 The illustrations herein show all the Motor Vehicle violations for
  • Category 1 of the violation types are those violations which do not require any information about the surroundings of the vehicle. Violations of the sort are determined only through the inspection of the vehicle’s state.
  • (A) shows vehicle 1, 3 driving without its head light beams ON 3 at incoming cars. This is the lawful use of headlights at night, because otherwise it would impede the visibility of the incoming cars by the glare of head-lights at full beam.
  • An artificial intelligence system composed of a CNN can ascertain the full beam condition against a non-full beam condition by training with enough sample sets, resulting in a binary format; 6 shows the correct use of Full beam. The 2 nd part of it would be to determine the license plate of the vehicle and then notifying the person of the said violation to teach them against doing it again.
  • Figure 2 A. shows examples of side mirrors which need to be deployed open while driving. Absence of the same is a violation and can be picked up by training the Artificial Neural Networks (ANNs) to detect side mirrors on vehicles.
  • ANNs Artificial Neural Networks
  • ANNs can detect the type of vehicle and the number of people traveling on the vehicle by identifying persons within the ROI of the vehicle.
  • ANNs can additionally count the total number of animals in the ROI can calculate the space of the ROI (vehicle container) then flag the vehicle if the space isn’t enough.
  • H. not wearing a seat belt is flagged is a seat belt is not recognized by the CNN / ANN.
  • a goods vehicle is identified by the ANNs and animals are similarly identified and any overlap in the bounding boxes is flagged as a violation by the driver of the vehicle.
  • Figure 3 A. without adequate lights 39, showing the license plates forces another violation trigger. B. instead of being offensive, some number plates are just generally in the wrong format, but not offensive. Wherein the size of the font of the angle of the font (italics etc.) or just the spacing between letter and number isn’t standard. C. for transportation vehicles it is not authorized to carry more than 6 persons on the same vehicle. In such cases, the ANNs can count the number of people riding and the type of vehicle and flag the vehicle for violating the rule. D. driving by a type of person who is unable to properly sit on the vehicle is detected by the ANNs and flagged. E.
  • ANNs can detect conditions when a mobile phone is being operated by observing the location of the hands, the visuals of the mobile phone.
  • Figure 4 A. Non-present wipers can be easily picked up by ANNs as well, similar to B for pollution creating vehicles; C. Safety devices are mandatory for motorcycles and scooters and identification of the missing components when trained against Proper vehicles can flag the ANN’s citation of the same. D. Good carriers need to display their credentials and also various other markings indicating what they are carrying within as well. ANNs can read the said signage and make sure it’s properly printed around the transport vehicles to comply with the local laws. E. Often trucks and goods carriers, due to their delays and/or intent to transport more than regulated in a single trip, may increase their carriage and introduce more goods and loads, thus increasing the overall height of the load, endangering other vehicles and breaching safety regulations.
  • ANNs can be trained against dimensions of a compliant truck load and non-compliant truck load.
  • F where civilians and drivers may use their vehicle to carry goods in contravention of the laws.
  • Passenger vehicles are not designed to carry goods and have to be embodied with out of scope fittings, etc. to ensure goods can be hosted, in contravention of passenger vehicle laws.
  • Artificial Neural Networks can detect the same and cite the vehicle for misuse of the vehicle.
  • Dangerous projection of goods G is a gross offence that costs both life and damage to property in various situations; we can train the ANNs to detect vehicles which carry such protruding loads / rods etc. H.
  • an ANN can detect whether such is visible or not through training. I, require goods vehicles to be fitted with the appropriate lights, especially when traveling in low light conditions to signal the approach and proximity of such vehicles to others. Missing one or more of the lights can be signaled as a non-compliance by having the ANNs train to detect the lights and consequently lack thereof.
  • Figure 5 Vehicles are not allowed to drive without the proper reflectors for low light conditions A; vehicles are also not allowed to carry loads beyond their bounding dimensions B, D; vehicles are not allowed to use their tailboard to carry loads on the vehicles. All load should be within the body of an open trailer / truck and the tailboard needs to be closed (not open and suspended to allow more load to be fitted in) C. Carrying goods in public service vehicle such as auto-rikshaws, tempos, e-rikshaws etc. is strictly prohibited and is recognized by training with data showing the same. E.
  • Figure 6 A. Transport vehicles or even non-transport vehicles carrying unsecured loads can be detected by the means of movement or displacement of the load wrf to the constraints of the body of the vehicle, C, Transport vehicles may not advertise without permission from the authorities. License plates need to conform to Transport and Non- Transport vehicles. F. Destination board should be illuminated for public carriers and can be picked up by ANNs. Similarly overloading vehicles with passengers or goods is a contravention of local laws in many cities G, H, I.; B, E again show the proper license plates for the vehicles.
  • Figure 7 A indicates traffic violations when a driver of a vehicle is not driving in proper lane.
  • B indicates traffic violation when the driver is driving in the center of the road creating inconvenience to other vehicles on the road and impedes overtaking etc.
  • C indicates the traffic violation when the driver stops the vehicle 95 on pedestrian crossing 93 (zebra crossing).
  • D indicates traffic violation when the driver 97, 98 is driving his vehicle on footpath (walking path for pedestrians) 96.
  • E indicates traffic violation driver overtakes any vehicle passing through an opposite lane of a road and cross the double yellow line.
  • F indicates traffic violation, when the driver stops in the opposite lane on any signal 102.
  • G indicates traffic violation when the driver of the vehicle l05does not stop even when the signal is red 106 (indicating stop for vehicle; a red-light).
  • H-I indicate various driving signs and not following such signs leading to a traffic violation.
  • Figure 8. indicates traffic violations when a driver of a vehicle 109 is driving on a one way street/road. B indicates traffic violation when the driver drives at times or locations when it is unauthorized. C indicates the traffic violation when the driver parks vehicle on footpath (walking path for pedestrians). D indicates traffic violation when the driver is driving his vehicle in spite of seeing/having seen a stopping order by the authorities. E indicates traffic violation to cause to allow a motor vehicle to be left in a position likely to cause undue inconvenience to other road users.. F indicates traffic violation when the driver leaves the vehicle in an unsafe position causing hindrance. G indicates traffic violation when the driver of a Bus stops without a Bus Stop . H indicates traffic violation when the driver parks the vehicle on pedestrian crossing (zebra crossing).
  • FIG. 9 A indicates traffic violations when a driver of a vehicle stops/parks his vehicle when no parking sign prohibits vehicles parked in that area.
  • B indicates traffic violations when a vehicle 128 stops/parks in front of any gate 129.
  • C indicates the traffic violation when the driver parks vehicle in such a manner that it may create inconvenience to the other vehicle for parking, i.e. not in any prescribed manner.
  • D indicates traffic violation when the driver 131, 132, 133 parked causing obstruction.
  • E indicates traffic violation when the vehicle is parked away from the footpath/sidewalk.
  • F indicates traffic violation when the driver parks on a bridge or flyover, or overpass.
  • G indicates traffic violation when the driver of the vehicle parks his vehicle in an authorized vehicle parking spot such as a Taxi Stand.
  • H indicates traffic violation when the driver stops/parks the vehicle 142 at a traffic island 144.
  • I indicate traffic violation when distance vehicle is parked within 15 meters of a bus stop.
  • FIG. 10 A indicates traffic violations when a driver of a does not allow other vehicles to overtake.
  • B indicates traffic violations when a driver of a vehicle takes a turn without giving signal indication.
  • C-E indicate the traffic violation when a vehicle is towed/pulled by another vehicle.
  • F indicates traffic violation when the driver tries to overtake or drive his vehicle in a rough/rash/negligent driving fashion, causing others danger and inconvenience.
  • G indicates traffic violation when the driver tries to drive his vehicle in dangerous or hasty driving manner, cutting through lanes and not driving in any orderly manner, breaking multiple rules of speed, overtaking etc..
  • H indicates traffic violation when the driver of the vehicle exceeds the speed limit provided on a road. I illustrates a picture of obeying traffic rules, i.e., when a security official asks to stop your vehicle, stop it.
  • FIG 11 A indicates racing on the roads. B overtaking perilously / dangerously. C overtaking from wrong side, indicates the traffic violation when a driver of a vehicle overtakes from left hand side. D indicates the traffic violation when a driver of a vehicle reverses his vehicle without care and attention. E indicates the traffic violation when a driver of a vehicle does not slow down at an intersection. F-H indicates indicate a picture of dis-obeying traffic rules, i.e., when a police officer asks to stop your vehicle, stop it. I indicate the correct way to approach another car on the freeway, it is a traffic violation when a driver of a vehicle flashes light when not required or dangerous.
  • Figure 12 A shows the installation of a clicker or a gesture recorder 193 on the steering wheel 194 of a car which connects wirelessly (Ultrasonic 192, Infrared 191, Bluetooth 190, WiFi 189) through one or more communication standards; connects to the camera capture device or a mobile device 188 mounted 187 on the windshield or dashboard of the car.
  • B shows an embodiment of the gesture capture device 195 having two buttons 196 mounted using a bracket on the steering wheel 194.
  • C showing that the same gesture capture device is mounted on a different area of the steering wheel 197.
  • Figure D shows the clasp/bracket that mounts the clicker or gesture capture device to the steering wheel.
  • FIG 13 illustrate exemplary various mountings for the gesture control device or the clicker.
  • a dedicated device may be provided on a steering wheel of a vehicle which enables the driver to communicate with the other remote devices avoiding any disturbance/diversions from driving his vehicle and can drive and capture violation through the click of a button hands-free from the mobile camera device.
  • the trigger initiated is delivered to the mobile camera device or the data collector device wirelessly.
  • A shows the same gesture capture device on a string on a person’s hand 201.
  • Figure 14 Shows the gesture capture device could be capacitive touch in nature as well and be mounted using an electrostatic or magnetic attachment to a part of a person’s body or to the vehicle directly.
  • Figure 14 A the secondary camera on the phone (or any dual camera device) can capture the movement of head and eyes 208 of a person driving the vehicle and match them to that being captured from the primary camera of the mobile camera device 212. The location of the advertising boards, traffic signs and other displays on the road are also monitored to decide the placement of them. The AI system further determines if the visual markers are indeed watched by people, can rate their quality of placement or determine if they are distracting and causing accidents / bad driving. This will help in deciding the placement of visual markers and changing them to avoid further accidents. B.
  • Figures 15 illustrates an exemplary flow chart of a typical application and describes how a person can use a mobile application or display device to simultaneously act as a specialist and also help collect data for AI Training for automated Specialists in Figure 16 and Figure 19.
  • a user can launch the application on mobile or any other computing handheld device etc..
  • User On submitting username and password 227, User is on the home screen 228 where the user can start as specialist and start capturing violations viewed inside videos collected by the data collectors; various options are available to the Specialist for the purpose.
  • the next option 229 is used to crop and collect the picture of a vehicle, 232 determines by ANPR or manually 231 the license plate of the vehicle, upon approval, 230 it goes to the next step.
  • the user is taken to another screen to manually enter the license number 233. Any change in the user’s entry compared to the AI / ANPR determined license plates, results in the system learning the Human Input and adds that to the long term learning of the AI engine for improving its performance in the future.
  • the users then choose the type of the challans/violations 236, 237 and submit it to the server; the appropriate citation amount is listed 235 for the offences.
  • the AI server/system then used the video for either training or restoring it to cloud for challan/citation recording.
  • the AI server 241 (also expanded in Figure 16) uses the human inputs and adds them to the pool of data points for the next iterations of training the weights on the neural networks. Then the specialist is presented a new video 228 and the cycle repeats.
  • Figure 16 expands the Neural Network Specialists and auditors (AI server) describing how images are received with ROIs 245 and then the outputs are trained using convolution and sub- sampling 246 to ultimately determine one or more traffic violations 247.
  • AI server Neural Network Specialists and auditors
  • This is a general purpose Neural Network shown here.
  • Various trained data is then used for generating the various outputs for automated detection.
  • the AI Engine comes online with higher efficacy to act as the automated specialist.
  • the cycle continues and continually automates the process of improving the AI engine (inside the AI Servers 244).
  • Various SSDs can be used for the purpose, wherein a separate SSD us used for detecting characters in a license plate.
  • Figures 17A-D illustrates how the specialist can do their job on a mobile phone or any other device having network connectivity.
  • 250 is the manually captured image of a vehicle and the manual capture of license plate characters is shown in B.
  • D, 263 shows various icons to represent various violation categories, which are selected to add to the violation list for the vehicles.
  • Figure 18 illustrates various other objects of interest being categorized by specialists on their mobile or web devices 270, 276, 280, 282. Then these categorized objects of interest are sent to the AI server for training the AI Specialists for future automated detection.
  • FIG. 19 Shows the Roadway stewardship network as explained in the Patent application number. 15/689350 filed on August 29, 2017, titled“On-demand Roadway Stewardship System previously invented by the author.
  • the Roadway stewardship network shows the Data Collectors at the very top (Blimps 284, Drones 285, Portable Cams 286, Camera glasses 287, Satellites 288, Robots 289, Police 290, Cars 291, Scooters 292, Buses 293, Trucks 294, Bicycles 295, Auto Rikshaws 296, among others), who fed the videos or series of images to the Cloud network.
  • the Specialists 304 review the data and look for objects of interest and then mark them for learning for the AI Specialists.
  • the auditors review the marked-up data from the Cloud 2 and similarly approve or disapprove the data for future processing; in cases also editing the information to correct for any mistakes or missing information/data-points. Finally the outputs like citations, tickets, and other interested data sets are sent to the required agencies, authorities for further processing / executive decision making.
  • FIG. 20 Shows the PARKSAFE setup wherein the market place has many unauthorized parking of vehicles 311 taking place.
  • the License plate 312 detectors B capture the plates and inform the illegally parked vehicle owners via mobile phones or app notifications to fix their parking or get a citation. Additionally it shows available parking spaces to them 314. Drivers using their mobile app D 316 can see the available parking spaces and pay digitally to reserve them.
  • C shows the Parking owners’ interface where they can put their available parking spaces for rent 315.
  • FIG. 21 a module diagram of artificial intelligence (AI) based system (2100) and its overall working, in accordance with an exemplary embodiment of the present disclosure.
  • the system may include one or more processor(s) (2102), an input/output (I/O) interface(s) (2104) and a memory (2106) for example, non-transitory storage device.
  • the processor may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational instructions.
  • the at least one processor is configured to fetch and execute computer-readable instructions stored in the memory.
  • the I/O interface may include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like.
  • the I/O interface may allow system to interact with a user directly.
  • the I/O interface can facilitate multiple communications within a wide variety of networks and protocol types, including wired networks, for example, LAN, cable, etc., and wireless networks, such as WLAN, cellular, or satellite.
  • the I/O interface may include one or more ports for connecting a number of devices to one another or to another server.
  • the memory may include any computer-readable medium known in the art including, for example, volatile memory, such as static random access memory (SRAM) and dynamic random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes.
  • volatile memory such as static random access memory (SRAM) and dynamic random access memory (DRAM)
  • non-volatile memory such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes.
  • the memory may include modules and data.
  • the modules include routines, programs, and objects, components, and data structures etc., which perform particular tasks or implement particular abstract data types.
  • the modules may include a receiver module (2108) to receive at least an object detection signal from one or more vision sensing systems, a detector module (2110) (alternatively or interchangeably referred to as “a detection module (2116)”) to determine, for said objects, a region of interest (ROI) selected from the received images or a series of images or a video, a training module (2112) to receive the manually classified objects, obtained from the images or a series of images or a video and trains the Neural Network to improve the detector performance, a logic module (2114) to receive as input the detected objects of interest in a series of one or more images or videos and determines various actions or events of interest, an annotation module (2118) to read license plates detected by the detector module, an occlusion module (2120) to hide objects ROI or hides non-ROI information in the image or series of images or video, a calculation module (2122) to calculate a speed of said object utilizing object detection from the location of the said object in different frames of the video or series of images and the relative velocity
  • the proposed system has been elaborated as above to include all the main modules, it is completely possible that actual implementations may include only a part of the proposed modules or a combination of those or a division of those into sub-modules in various combinations across multiple devices that can be operatively coupled with each other, including in the cloud. Further the modules can be configured in any sequence to achieve objectives elaborated. Also, it can be appreciated that proposed system can be configured in a computing device or across a plurality of computing devices operatively connected with each other, wherein the computing devices can be any of a computer, a laptop, a smartphone, an Internet enabled mobile device and the like. All such modifications and embodiments are completely within the scope of the present disclosure.
  • the term“coupled to” is intended to include both direct coupling (in which two elements that are coupled to each other or in contact each other) and indirect coupling (in which at least one additional element is located between the two elements). Therefore, the terms “coupled to” and “coupled with” are used synonymously. Within the context of this document terms“coupled to” and“coupled with” are also used euphemistically to mean“communicatively coupled with” over a network, where two or more devices are able to exchange data with each other over the network, possibly via one or more intermediary device. [0203] Moreover, in interpreting both the specification and the claims, all terms should be interpreted in the broadest possible manner consistent with the context.
  • the terms“comprises” and“comprising” should be interpreted as referring to elements, components, or steps in a non-exclusive manner, indicating that the referenced elements, components, or steps may be present, or utilized, or combined with other elements, components, or steps that are not expressly referenced.
  • the specification claims refers to at least one of something selected from the group consisting of A, B, C ....and N, the text should be interpreted as requiring only one element from the group, not A plus N, or B plus N, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Human Computer Interaction (AREA)
  • Mathematical Physics (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Ophthalmology & Optometry (AREA)
  • Databases & Information Systems (AREA)
  • Psychiatry (AREA)
  • Social Psychology (AREA)
  • Medical Informatics (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Traffic Control Systems (AREA)
  • Image Analysis (AREA)
  • Document Processing Apparatus (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

La présente invention concerne des systèmes basés sur l'intelligence artificielle et un procédé de détermination d'infractions à la circulation. Cette invention utilise des réseaux neuronaux convolutifs profonds et des algorithmes basés sur la vision artificielle pour effectuer une tâche de détection et de reconnaissance afin de fournir une solution complète assurant aux usagers des routes un stationnement, une conduite et des trajets sûrs, respectueux de la loi et agréables. Les systèmes d'administration de routes et les systèmes de gestion de stationnement, lorsqu'ils sont à la demande et communautaires, peuvent jouer un rôle très important dans la régulation des conditions de conduite dans les villes et sur les autoroutes. Grâce à l'automatisation d'un système d'administration de routes à la demande communautaire au moyen de sous-systèmes d'intelligence artificielle (IA), les utilisateurs peuvent être habitués à reconnaître les lois et réglementations quant à l'usage des routes et être informés de celles-ci. Le processus peut être facilité par une console interactive/un jeu, ce qui peut également servir à la monétisation afin que les individus gagnent de l'argent en échange de leur contribution.
PCT/IB2019/051294 2018-03-12 2019-02-18 Système d'intelligence artificielle et d'administration de routes à la demande Ceased WO2019175686A1 (fr)

Priority Applications (5)

Application Number Priority Date Filing Date Title
AU2019235551A AU2019235551B2 (en) 2018-03-12 2019-02-18 On-demand artificial intelligence and roadway stewardship system
EP19766866.8A EP3676754A4 (fr) 2018-03-12 2019-02-18 Système d'intelligence artificielle et d'administration de routes à la demande
SG11201909815R SG11201909815RA (en) 2018-03-12 2019-02-18 On-demand artificial intelligence and roadway stewardship system
CA3093836A CA3093836A1 (fr) 2018-03-12 2019-02-18 Systeme d'intelligence artificielle et d'administration de routes a la demande
AU2021254525A AU2021254525A1 (en) 2018-03-12 2021-10-18 On-demand artificial intelligence and roadway stewardship system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US15/919,033 US10296794B2 (en) 2016-12-20 2018-03-12 On-demand artificial intelligence and roadway stewardship system
US15/919,033 2018-03-12

Publications (1)

Publication Number Publication Date
WO2019175686A1 true WO2019175686A1 (fr) 2019-09-19

Family

ID=67908733

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2019/051294 Ceased WO2019175686A1 (fr) 2018-03-12 2019-02-18 Système d'intelligence artificielle et d'administration de routes à la demande

Country Status (5)

Country Link
EP (1) EP3676754A4 (fr)
AU (2) AU2019235551B2 (fr)
CA (1) CA3093836A1 (fr)
SG (1) SG11201909815RA (fr)
WO (1) WO2019175686A1 (fr)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807429A (zh) * 2019-10-23 2020-02-18 西安科技大学 基于tiny-YOLOv3的施工安全检测方法及系统
CN110909626A (zh) * 2019-11-04 2020-03-24 上海眼控科技股份有限公司 车辆压线检测方法、装置、移动终端及存储介质
CN111259796A (zh) * 2020-01-16 2020-06-09 东华大学 一种基于图像几何特征的车道线检测方法
CN111523544A (zh) * 2020-04-23 2020-08-11 上海眼控科技股份有限公司 车牌类型检测方法、系统、计算机设备及可读存储介质
CN111523488A (zh) * 2020-04-26 2020-08-11 上海集光安防科技股份有限公司 厨房工作人员行为实时监控方法
CN111582189A (zh) * 2020-05-11 2020-08-25 腾讯科技（深圳）有限公司 交通信号灯识别方法、装置、车载控制终端及机动车
CN111814762A (zh) * 2020-08-24 2020-10-23 深延科技(北京)有限公司 头盔佩戴检测方法和装置
CN113269005A (zh) * 2020-02-14 2021-08-17 深圳云天励飞技术有限公司 一种安全带检测方法、装置及电子设备
CN113486764A (zh) * 2021-06-30 2021-10-08 中南大学 一种基于改进的YOLOv3的坑洼检测方法
US20220042817A1 (en) * 2020-08-04 2022-02-10 Toyota Research Institute, Inc. Systems and methods for map verification
CN114092897A (zh) * 2020-07-30 2022-02-25 罗伯特·博世有限公司 道路分类方法与设备
CN117423093A (zh) * 2023-12-18 2024-01-19 深圳市深航华创汽车科技有限公司 基于行车记录仪的行车检测方法、装置、设备及存储介质
EP4139177A4 (fr) * 2020-04-21 2024-05-01 Micron Technology, Inc. Contrôle de passagers
CN118379671A (zh) * 2024-06-27 2024-07-23 四川国蓝中天环境科技集团有限公司 一种基于视频和路径规划的城市道路破损识别方法
EP4445307A4 (fr) * 2021-12-07 2024-11-27 Biosphera Inc. Procédé de quantification du niveau de performance d'une activité humaine dirigée contre l'atténuation d'effets anthropogènes sur l'environnement

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111380545B (zh) * 2015-02-10 2024-11-12 御眼视觉技术有限公司 用于自主车辆导航的方法、服务器、自主车辆以及介质
US11281917B2 (en) * 2019-10-31 2022-03-22 Aptiv Technologies Limited Multi-domain neighborhood embedding and weighting of point cloud data
US12230139B2 (en) * 2021-11-18 2025-02-18 Here Global B.V. Predicting traffic violation hotspots using map features and sensors data

Citations (64)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5448484A (en) * 1992-11-03 1995-09-05 Bullock; Darcy M. Neural network-based vehicle detection system and method
US6546119B2 (en) 1998-02-24 2003-04-08 Redflex Traffic Systems Automated traffic violation monitoring and reporting system
US6560529B1 (en) 1998-09-15 2003-05-06 Robert Bosch Gmbh Method and device for traffic sign recognition and navigation
EP1096457B1 (fr) 1999-10-29 2004-02-25 Volkswagen Aktiengesellschaft Procédé et dispositif de reconaissance électronique de panneaux de signalisation routière
WO2007023103A1 (fr) 2005-08-23 2007-03-01 Siemens Aktiengesellschaft Systeme pour fournir un avertissement en cas de sortie de la voie et/ou pour offrir une fonction de maintien dans la voie
US7363133B2 (en) 2005-03-21 2008-04-22 Valeo Schalter Und Sensoren Gmbh Lane detection system and method
CN101183427A (zh) 2007-12-05 2008-05-21 浙江工业大学 基于计算机视觉的违章停车检测装置
US7720580B2 (en) 2004-12-23 2010-05-18 Donnelly Corporation Object detection system for vehicle
US20110109476A1 (en) 2009-03-31 2011-05-12 Porikli Fatih M Method for Recognizing Traffic Signs
US8184863B2 (en) 2007-01-05 2012-05-22 American Traffic Solutions, Inc. Video speed detection system
US20120148105A1 (en) 2010-12-14 2012-06-14 Xerox Corporation Automated license plate recognition system and method using human-in-the-loop based adaptive learning
EP2487454A2 (fr) 2011-02-14 2012-08-15 Scania CV AB Estimation de longueur de véhicule
CN202422420U (zh) 2012-01-13 2012-09-05 大连民族学院 一种基于视频监控的违章停车检测系统
WO2013009697A1 (fr) 2011-07-08 2013-01-17 Bendix Commercial Vehicle Systems Llc Procédé et appareil de mesure de distance et de détection de véhicule basées sur une image
US20130049988A1 (en) 2011-08-24 2013-02-28 Ford Global Technologies, Llc Device and method for traffic sign recognition
EP2574958A1 (fr) 2011-09-28 2013-04-03 Honda Research Institute Europe GmbH Procédé de détection de terrain sur route et système pour systèmes d'assistance au conducteur
EP2578464A1 (fr) 2011-10-06 2013-04-10 Honda Research Institute Europe GmbH Système d'alerte de type vidéo pour un véhicule
US8446467B2 (en) 2006-05-05 2013-05-21 Scott K. Tilton Combined speed detection, video and timing apparatus
US8509486B2 (en) 2010-10-29 2013-08-13 National Chiao Tung University Vehicle license plate recognition method and system thereof
US8587649B2 (en) 2009-04-21 2013-11-19 Create Electronic Optical Co., Ltd. Lane departure warning system
US20130311075A1 (en) 2012-05-18 2013-11-21 Continental Automotive Systems, Inc. Motorcycle and helmet providing advance driver assistance
WO2014130178A1 (fr) 2013-02-21 2014-08-28 Google Inc. Procédé pour la détection de conducteurs agressifs situés à proximité et l'ajustement des modes de conduite
CN104200668A (zh) 2014-07-28 2014-12-10 四川大学 一种基于图像分析的摩托车不戴头盔违章事件检测方法
CN104298976A (zh) 2014-10-16 2015-01-21 电子科技大学 基于卷积神经网络的车牌检测方法
US8983136B2 (en) 2011-09-23 2015-03-17 Ricoh Company, Ltd. Traffic sign detecting method and traffic sign detecting device
EP2863338A2 (fr) 2013-10-16 2015-04-22 Xerox Corporation Identification de véhicule retardée pour mise en place de la confidentialité
WO2015056105A1 (fr) 2013-10-14 2015-04-23 Mobileye Vision Technologies Ltd. Système d'imagerie multiple orientée vers l'avant pour la navigation d'un véhicule
US20150235091A1 (en) 2014-02-14 2015-08-20 Denso Corporation Lane-line recognition apparatus
EP2966590A1 (fr) 2014-07-07 2016-01-13 HERE Global B.V. Trafic de niveau de voie
US9286524B1 (en) 2015-04-15 2016-03-15 Toyota Motor Engineering & Manufacturing North America, Inc. Multi-task deep convolutional neural networks for efficient and robust traffic lane detection
US9305223B1 (en) 2013-06-26 2016-04-05 Google Inc. Vision-based indicator signal detection using spatiotemporal filtering
WO2016069253A1 (fr) 2014-10-28 2016-05-06 Trw Automotive U.S. Llc Détection de file augmentée faisant appel a des données cinématiques
EP3038012A1 (fr) 2013-07-17 2016-06-29 Kaparazoom SLU Signal d'identification de feux de circulation pour vision informatisée
US9428192B2 (en) 2004-04-15 2016-08-30 Magna Electronics Inc. Vision system for vehicle
CN105975929A (zh) 2016-05-04 2016-09-28 北京大学深圳研究生院 一种基于聚合通道特征的快速行人检测方法
US20160293002A1 (en) 2013-05-27 2016-10-06 Ekin Teknoloji Sanayi Ve Ticaret Anonim Sirketi Mobile number plate recognition and speed detection system
CN106022300A (zh) 2016-06-02 2016-10-12 中国科学院信息工程研究所 基于级联深度学习的交通标志识别方法和系统
CN106250892A (zh) 2016-08-11 2016-12-21 万永秀 车牌字符自动识别方法
CN106295551A (zh) 2016-08-05 2017-01-04 南京理工大学 一种基于视频分析的人员安全帽佩戴情况实时检测方法
CN106327878A (zh) 2016-11-16 2017-01-11 天津市中环系统工程有限责任公司 一种移动违停抓拍系统及实现方法
US20170017848A1 (en) 2015-07-17 2017-01-19 Magna Electronics Inc. Vehicle parking assist system with vision-based parking space detection
CN106373426A (zh) 2016-09-29 2017-02-01 成都通甲优博科技有限责任公司 基于计算机视觉的停车位及违规占道停车监控方法
CN106372662A (zh) 2016-08-30 2017-02-01 腾讯科技（深圳）有限公司 安全帽佩戴的检测方法和装置、摄像头、服务器
CN106408000A (zh) 2016-08-25 2017-02-15 国网浙江省电力公司杭州供电公司 一种人员安全帽智能检测方法及装置
EP3143607A1 (fr) 2014-05-14 2017-03-22 Mobileye Vision Technologies Ltd. Systèmes et procédés de détection de bordure de trottoir et d'évaluation des risques pour les piétons
CN106652465A (zh) 2016-11-15 2017-05-10 成都通甲优博科技有限责任公司 一种道路异常驾驶行为的识别方法及系统
CN106845430A (zh) 2017-02-06 2017-06-13 东华大学 基于加速区域卷积神经网络的行人检测与跟踪方法
CN106845480A (zh) 2017-01-13 2017-06-13 河海大学 一种从图片中识别车牌的方法
CN106874863A (zh) * 2017-01-24 2017-06-20 南京大学 基于深度卷积神经网络的车辆违停逆行检测方法
CN106874907A (zh) 2017-01-19 2017-06-20 博康智能信息技术有限公司北京海淀分公司 一种建立车牌识别模型的方法及装置
CN106935035A (zh) 2017-04-07 2017-07-07 西安电子科技大学 基于ssd神经网络的违章停车车辆实时检测方法
US9704060B2 (en) 2013-12-17 2017-07-11 Cloud Computing Center Chinese Academy Of Sciences Method for detecting traffic violation
WO2017136578A1 (fr) 2016-02-03 2017-08-10 Honda Motor Co., Ltd. Détection d'objets partiellement occlus au moyen d'un classement de contexte et de profondeur
CN107154045A (zh) 2017-03-31 2017-09-12 南京航空航天大学 一种基于计算机视觉的导流区违章轧线车辆精确定位方法
US9760806B1 (en) 2016-05-11 2017-09-12 TCL Research America Inc. Method and system for vision-centric deep-learning-based road situation analysis
KR20170105845A (ko) 2016-03-10 2017-09-20 현대오토에버 주식회사 운전지원장치, 거리감지방법 및 차로의 너비감지방법
US9779314B1 (en) 2014-08-21 2017-10-03 Waymo Llc Vision-based detection and classification of traffic lights
US20170300763A1 (en) 2016-04-19 2017-10-19 GM Global Technology Operations LLC Road feature detection using a vehicle camera system
CN107481526A (zh) 2017-09-07 2017-12-15 公安部第三研究所 用于行车变道检测记录及违章变道举报控制的系统及方法
CN107491753A (zh) 2017-08-16 2017-12-19 电子科技大学 一种基于背景建模的违章停车检测方法
US9852019B2 (en) * 2013-07-01 2017-12-26 Agent Video Intelligence Ltd. System and method for abnormality detection
EP3261017A1 (fr) 2016-06-20 2017-12-27 Delphi Technologies, Inc. Système de traitement d'image pour détecter des objets d'intérêt
US20170372161A1 (en) 2016-06-24 2017-12-28 Accenture Global Solutions Limited Intelligent automatic license plate recognition for electronic tolling environments
CN107730906A (zh) 2017-07-11 2018-02-23 银江股份有限公司 斑马线车辆不礼让行人行为的视觉检测系统

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9477892B2 (en) * 2014-03-26 2016-10-25 Xerox Corporation Efficient method of offline training a special-type parked vehicle detector for video-based on-street parking occupancy detection systems

Patent Citations (65)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5448484A (en) * 1992-11-03 1995-09-05 Bullock; Darcy M. Neural network-based vehicle detection system and method
US6546119B2 (en) 1998-02-24 2003-04-08 Redflex Traffic Systems Automated traffic violation monitoring and reporting system
US6560529B1 (en) 1998-09-15 2003-05-06 Robert Bosch Gmbh Method and device for traffic sign recognition and navigation
EP1096457B1 (fr) 1999-10-29 2004-02-25 Volkswagen Aktiengesellschaft Procédé et dispositif de reconaissance électronique de panneaux de signalisation routière
US9428192B2 (en) 2004-04-15 2016-08-30 Magna Electronics Inc. Vision system for vehicle
US7720580B2 (en) 2004-12-23 2010-05-18 Donnelly Corporation Object detection system for vehicle
US7363133B2 (en) 2005-03-21 2008-04-22 Valeo Schalter Und Sensoren Gmbh Lane detection system and method
WO2007023103A1 (fr) 2005-08-23 2007-03-01 Siemens Aktiengesellschaft Systeme pour fournir un avertissement en cas de sortie de la voie et/ou pour offrir une fonction de maintien dans la voie
US8446467B2 (en) 2006-05-05 2013-05-21 Scott K. Tilton Combined speed detection, video and timing apparatus
US8184863B2 (en) 2007-01-05 2012-05-22 American Traffic Solutions, Inc. Video speed detection system
CN101183427A (zh) 2007-12-05 2008-05-21 浙江工业大学 基于计算机视觉的违章停车检测装置
US20110109476A1 (en) 2009-03-31 2011-05-12 Porikli Fatih M Method for Recognizing Traffic Signs
US8587649B2 (en) 2009-04-21 2013-11-19 Create Electronic Optical Co., Ltd. Lane departure warning system
US8509486B2 (en) 2010-10-29 2013-08-13 National Chiao Tung University Vehicle license plate recognition method and system thereof
US20120148105A1 (en) 2010-12-14 2012-06-14 Xerox Corporation Automated license plate recognition system and method using human-in-the-loop based adaptive learning
EP2487454A2 (fr) 2011-02-14 2012-08-15 Scania CV AB Estimation de longueur de véhicule
WO2013009697A1 (fr) 2011-07-08 2013-01-17 Bendix Commercial Vehicle Systems Llc Procédé et appareil de mesure de distance et de détection de véhicule basées sur une image
US20130049988A1 (en) 2011-08-24 2013-02-28 Ford Global Technologies, Llc Device and method for traffic sign recognition
US8983136B2 (en) 2011-09-23 2015-03-17 Ricoh Company, Ltd. Traffic sign detecting method and traffic sign detecting device
EP2574958A1 (fr) 2011-09-28 2013-04-03 Honda Research Institute Europe GmbH Procédé de détection de terrain sur route et système pour systèmes d'assistance au conducteur
EP2578464A1 (fr) 2011-10-06 2013-04-10 Honda Research Institute Europe GmbH Système d'alerte de type vidéo pour un véhicule
CN202422420U (zh) 2012-01-13 2012-09-05 大连民族学院 一种基于视频监控的违章停车检测系统
US20130311075A1 (en) 2012-05-18 2013-11-21 Continental Automotive Systems, Inc. Motorcycle and helmet providing advance driver assistance
WO2014130178A1 (fr) 2013-02-21 2014-08-28 Google Inc. Procédé pour la détection de conducteurs agressifs situés à proximité et l'ajustement des modes de conduite
US20160293002A1 (en) 2013-05-27 2016-10-06 Ekin Teknoloji Sanayi Ve Ticaret Anonim Sirketi Mobile number plate recognition and speed detection system
US9305223B1 (en) 2013-06-26 2016-04-05 Google Inc. Vision-based indicator signal detection using spatiotemporal filtering
US9852019B2 (en) * 2013-07-01 2017-12-26 Agent Video Intelligence Ltd. System and method for abnormality detection
EP3038012A1 (fr) 2013-07-17 2016-06-29 Kaparazoom SLU Signal d'identification de feux de circulation pour vision informatisée
WO2015056105A1 (fr) 2013-10-14 2015-04-23 Mobileye Vision Technologies Ltd. Système d'imagerie multiple orientée vers l'avant pour la navigation d'un véhicule
EP2863338A2 (fr) 2013-10-16 2015-04-22 Xerox Corporation Identification de véhicule retardée pour mise en place de la confidentialité
US9704060B2 (en) 2013-12-17 2017-07-11 Cloud Computing Center Chinese Academy Of Sciences Method for detecting traffic violation
US20150235091A1 (en) 2014-02-14 2015-08-20 Denso Corporation Lane-line recognition apparatus
EP3143607A1 (fr) 2014-05-14 2017-03-22 Mobileye Vision Technologies Ltd. Systèmes et procédés de détection de bordure de trottoir et d'évaluation des risques pour les piétons
EP2966590A1 (fr) 2014-07-07 2016-01-13 HERE Global B.V. Trafic de niveau de voie
CN104200668A (zh) 2014-07-28 2014-12-10 四川大学 一种基于图像分析的摩托车不戴头盔违章事件检测方法
US9779314B1 (en) 2014-08-21 2017-10-03 Waymo Llc Vision-based detection and classification of traffic lights
CN104298976A (zh) 2014-10-16 2015-01-21 电子科技大学 基于卷积神经网络的车牌检测方法
WO2016069253A1 (fr) 2014-10-28 2016-05-06 Trw Automotive U.S. Llc Détection de file augmentée faisant appel a des données cinématiques
US9286524B1 (en) 2015-04-15 2016-03-15 Toyota Motor Engineering & Manufacturing North America, Inc. Multi-task deep convolutional neural networks for efficient and robust traffic lane detection
US20170017848A1 (en) 2015-07-17 2017-01-19 Magna Electronics Inc. Vehicle parking assist system with vision-based parking space detection
WO2017136578A1 (fr) 2016-02-03 2017-08-10 Honda Motor Co., Ltd. Détection d'objets partiellement occlus au moyen d'un classement de contexte et de profondeur
KR20170105845A (ko) 2016-03-10 2017-09-20 현대오토에버 주식회사 운전지원장치, 거리감지방법 및 차로의 너비감지방법
US20170300763A1 (en) 2016-04-19 2017-10-19 GM Global Technology Operations LLC Road feature detection using a vehicle camera system
WO2017190574A1 (fr) 2016-05-04 2017-11-09 北京大学深圳研究生院 Procédé de détection rapide de piétons basé sur des caractéristiques de canal d'agrégation
CN105975929A (zh) 2016-05-04 2016-09-28 北京大学深圳研究生院 一种基于聚合通道特征的快速行人检测方法
US9760806B1 (en) 2016-05-11 2017-09-12 TCL Research America Inc. Method and system for vision-centric deep-learning-based road situation analysis
CN106022300A (zh) 2016-06-02 2016-10-12 中国科学院信息工程研究所 基于级联深度学习的交通标志识别方法和系统
EP3261017A1 (fr) 2016-06-20 2017-12-27 Delphi Technologies, Inc. Système de traitement d'image pour détecter des objets d'intérêt
US20170372161A1 (en) 2016-06-24 2017-12-28 Accenture Global Solutions Limited Intelligent automatic license plate recognition for electronic tolling environments
CN106295551A (zh) 2016-08-05 2017-01-04 南京理工大学 一种基于视频分析的人员安全帽佩戴情况实时检测方法
CN106250892A (zh) 2016-08-11 2016-12-21 万永秀 车牌字符自动识别方法
CN106408000A (zh) 2016-08-25 2017-02-15 国网浙江省电力公司杭州供电公司 一种人员安全帽智能检测方法及装置
CN106372662A (zh) 2016-08-30 2017-02-01 腾讯科技（深圳）有限公司 安全帽佩戴的检测方法和装置、摄像头、服务器
CN106373426A (zh) 2016-09-29 2017-02-01 成都通甲优博科技有限责任公司 基于计算机视觉的停车位及违规占道停车监控方法
CN106652465A (zh) 2016-11-15 2017-05-10 成都通甲优博科技有限责任公司 一种道路异常驾驶行为的识别方法及系统
CN106327878A (zh) 2016-11-16 2017-01-11 天津市中环系统工程有限责任公司 一种移动违停抓拍系统及实现方法
CN106845480A (zh) 2017-01-13 2017-06-13 河海大学 一种从图片中识别车牌的方法
CN106874907A (zh) 2017-01-19 2017-06-20 博康智能信息技术有限公司北京海淀分公司 一种建立车牌识别模型的方法及装置
CN106874863A (zh) * 2017-01-24 2017-06-20 南京大学 基于深度卷积神经网络的车辆违停逆行检测方法
CN106845430A (zh) 2017-02-06 2017-06-13 东华大学 基于加速区域卷积神经网络的行人检测与跟踪方法
CN107154045A (zh) 2017-03-31 2017-09-12 南京航空航天大学 一种基于计算机视觉的导流区违章轧线车辆精确定位方法
CN106935035A (zh) 2017-04-07 2017-07-07 西安电子科技大学 基于ssd神经网络的违章停车车辆实时检测方法
CN107730906A (zh) 2017-07-11 2018-02-23 银江股份有限公司 斑马线车辆不礼让行人行为的视觉检测系统
CN107491753A (zh) 2017-08-16 2017-12-19 电子科技大学 一种基于背景建模的违章停车检测方法
CN107481526A (zh) 2017-09-07 2017-12-15 公安部第三研究所 用于行车变道检测记录及违章变道举报控制的系统及方法

Non-Patent Citations (16)

* Cited by examiner, † Cited by third party
Title
"Intelligent Vehicles Symposium Proceedings", 2014, IEEE, article "Integrating motion and appearance for overtaking vehicle detection"
"Signal Jump Detection Process", INTERNATIONAL JOURNAL OF COMPUTER APPLICATIONS TECHNOLOGY AND RESEARCH, vol. 6, 2017, pages 101 - 105, ISSN: 2319-8656
ANGEL OVA, ANELIA ET AL., REAL-TIME PEDESTRIAN DETECTION WITH DEEP NETWORK CASCADES
C. VISHNU, DETECTION OF MOTORCYCLISTS WITHOUT HELMET IN VIDEOS USING CONVOLUTIONAL NEURAL NETWORK
GURGHIAN: "DeepLanes: End-To-End Lane Position Estimation Using Deep Neural Networks", CVPR WORKSHOPS, pages 38 - 45
HALOI, MRINAL ET AL., A ROBUST LANE DETECTION AND DEPARTURE WARNING SYSTEM
HUANG, SHIH-SHINH ET AL.: "Robotics and Automation, 2004 Proceedings ICRA'04 2004 IEEE International Conference", vol. 3, 2004, IEEE, article "On-board vision system for lane recognition and front-vehicle detection to enhance driver's awareness"
MRINAL HALOI, TRAFFIC SIGN CLASSIFICATION USING DEEP INCEPTION BASED CONVOLUTIONAL NETWORKS
SERMANET, TRAFFIC SIGN RECOGNITION WITH MULTI-SCALE CONVOLUTIONAL NETWORKS
SUTIKNO, CLASSIFICATION OF MOTORCYCLISTS NOT WEAR HELMET ON DIGITAL IMAGE WITH BACKPROPAGATION NEURAL NETWORK
SYED ZAIN MASOOD, LICENSE PLATE DETECTION AND RECOGNITION USING DEEPLY LEARNED CONVOLUTIONAL NEURAL NETWORKS
XIE, XUEMEI ET AL.: "proceedings of the 2017 International Conference on Deep Learning Technologies", 2017, ACM, article "Real-Time Illegal Parking Detection System Based on Deep Learning"
XUEMEI XIE, REAL-TIME ILLEGAL PARKING DETECTION SYSTEM BASED ON DEEP LEARNING
Y. WU, TRAFFIC SIGN DETECTION BASED ON CONVOLUTIONAL NEURAL NETWORKS
ZENG, XINGYU ET AL., DEEP LEARNING OF SCENE-SPECIFIC CLASSIFIER FOR PEDESTRIAN DETECTION
ZHU ET AL.: "Traffic-sign detection and classification in the wild", PROCEEDINGS OF THE IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, pages 2110 - 2118

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110807429B (zh) * 2019-10-23 2023-04-07 西安科技大学 基于tiny-YOLOv3的施工安全检测方法及系统
CN110807429A (zh) * 2019-10-23 2020-02-18 西安科技大学 基于tiny-YOLOv3的施工安全检测方法及系统
CN110909626A (zh) * 2019-11-04 2020-03-24 上海眼控科技股份有限公司 车辆压线检测方法、装置、移动终端及存储介质
CN111259796A (zh) * 2020-01-16 2020-06-09 东华大学 一种基于图像几何特征的车道线检测方法
CN113269005B (zh) * 2020-02-14 2024-06-11 深圳云天励飞技术有限公司 一种安全带检测方法、装置及电子设备
CN113269005A (zh) * 2020-02-14 2021-08-17 深圳云天励飞技术有限公司 一种安全带检测方法、装置及电子设备
EP4139177A4 (fr) * 2020-04-21 2024-05-01 Micron Technology, Inc. Contrôle de passagers
CN111523544A (zh) * 2020-04-23 2020-08-11 上海眼控科技股份有限公司 车牌类型检测方法、系统、计算机设备及可读存储介质
CN111523488A (zh) * 2020-04-26 2020-08-11 上海集光安防科技股份有限公司 厨房工作人员行为实时监控方法
CN111582189A (zh) * 2020-05-11 2020-08-25 腾讯科技（深圳）有限公司 交通信号灯识别方法、装置、车载控制终端及机动车
CN114092897A (zh) * 2020-07-30 2022-02-25 罗伯特·博世有限公司 道路分类方法与设备
US20220042817A1 (en) * 2020-08-04 2022-02-10 Toyota Research Institute, Inc. Systems and methods for map verification
US12241754B2 (en) * 2020-08-04 2025-03-04 Toyota Research Institute, Inc. Systems and methods for map verification
CN111814762A (zh) * 2020-08-24 2020-10-23 深延科技(北京)有限公司 头盔佩戴检测方法和装置
CN113486764B (zh) * 2021-06-30 2022-05-03 中南大学 一种基于改进的YOLOv3的坑洼检测方法
CN113486764A (zh) * 2021-06-30 2021-10-08 中南大学 一种基于改进的YOLOv3的坑洼检测方法
EP4445307A4 (fr) * 2021-12-07 2024-11-27 Biosphera Inc. Procédé de quantification du niveau de performance d'une activité humaine dirigée contre l'atténuation d'effets anthropogènes sur l'environnement
CN117423093A (zh) * 2023-12-18 2024-01-19 深圳市深航华创汽车科技有限公司 基于行车记录仪的行车检测方法、装置、设备及存储介质
CN117423093B (zh) * 2023-12-18 2024-03-29 深圳市深航华创汽车科技有限公司 基于行车记录仪的行车检测方法、装置、设备及存储介质
CN118379671A (zh) * 2024-06-27 2024-07-23 四川国蓝中天环境科技集团有限公司 一种基于视频和路径规划的城市道路破损识别方法

Also Published As

Publication number Publication date
EP3676754A4 (fr) 2021-09-01
CA3093836A1 (fr) 2019-09-19
SG11201909815RA (en) 2019-11-28
EP3676754A1 (fr) 2020-07-08
AU2019235551B2 (en) 2022-02-03
AU2019235551A1 (en) 2020-09-24
AU2021254525A1 (en) 2021-11-11

Similar Documents

Publication Publication Date Title
US10296794B2 (en) On-demand artificial intelligence and roadway stewardship system
AU2019235551B2 (en) On-demand artificial intelligence and roadway stewardship system
Loce et al. Computer vision in roadway transportation systems: a survey
EP3416153B1 (fr) Procédé et système d'identification de place de parking
US20150377634A1 (en) Automatic content analysis method and system
Loce et al. Computer vision and imaging in intelligent transportation systems
KR102526583B1 (ko) 딥러닝 기반 교차로 객체검지 및 자동 알람 시스템
Ravish et al. Intelligent traffic violation detection
JP6524846B2 (ja) 車両特定装置およびこれを備えた車両特定システム
KR101394201B1 (ko) 버스장착 무선cctv 촬영단속시스템
Kumar et al. E-challan automation for RTO using OCR
WO2017022269A1 (fr) Dispositif d'identification de véhicule et système d'identification de véhicule le comprenant
JP6704568B1 (ja) 情報処理システム、情報処理装置、端末装置、サーバ装置、プログラム、又は方法
Songire et al. Using YOLO V7 development of Complete VIDS solution based on latest requirements to provide highway traffic and incident real time info to the ATMS control room using Artificial intelligence
Rithish et al. Evaluating computer vision techniques for urban mobility on large-scale, unconstrained roads
Kuo et al. Illegal parking detection based on multi-task driving perception
Ameer et al. Traffic violation detection system
Li et al. Smartphone-based method for automated speed enforcement
Mehboob et al. Mathematical model based traffic violations identification
Goel et al. Enforcement of Automatic Penalty (e-Penalty) to Govern the Traffic Rule Violators in Digitized INDIA Using ICT
Gowtham et al. An investigation approach used for pattern classification and recognition of an emergency vehicle
Li et al. Smartphone-based method for speed limit enforcement
Graham Electronic Toll Collections and Smart City Payments
KR102892590B1 (ko) 딥 러닝 기반의 지능형 영상 분석 시스템
Abraham et al. Computer Vision in Smart Parking Solution Systems: Enhancing Urban Mobility

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19766866

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019766866

Country of ref document: EP

Effective date: 20200330

ENP Entry into the national phase

Ref document number: 3093836

Country of ref document: CA

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019235551

Country of ref document: AU

Date of ref document: 20190218

Kind code of ref document: A