WO2025060413A1 - 行车盲区监测方法、装置、设备及可读存储介质 - Google Patents

行车盲区监测方法、装置、设备及可读存储介质 Download PDF

Info

Publication number
WO2025060413A1
WO2025060413A1 PCT/CN2024/088761 CN2024088761W WO2025060413A1 WO 2025060413 A1 WO2025060413 A1 WO 2025060413A1 CN 2024088761 W CN2024088761 W CN 2024088761W WO 2025060413 A1 WO2025060413 A1 WO 2025060413A1
Authority
WO
WIPO (PCT)
Prior art keywords
grayscale
images
value
frames
obstacle
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.)
Pending
Application number
PCT/CN2024/088761
Other languages
English (en)
French (fr)
Inventor
许鑫
李洋
吴鹏
方家萌
毛竹君
刘杏
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.)
Dongfeng Commercial Vehicle Co Ltd
Original Assignee
Dongfeng Commercial Vehicle Co Ltd
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
Application filed by Dongfeng Commercial Vehicle Co Ltd filed Critical Dongfeng Commercial Vehicle Co Ltd
Priority to EP24785958.0A priority Critical patent/EP4553794A4/en
Publication of WO2025060413A1 publication Critical patent/WO2025060413A1/zh
Anticipated expiration legal-status Critical
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/16—Anti-collision systems
    • G08G1/167—Driving aids for lane monitoring, lane changing, e.g. blind spot detection
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R1/00—Optical viewing arrangements; Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles
    • B60R1/20—Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles
    • B60R1/22—Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles for viewing an area outside the vehicle, e.g. the exterior of the vehicle
    • B60R1/23—Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles for viewing an area outside the vehicle, e.g. the exterior of the vehicle with a predetermined field of view
    • B60R1/25—Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles for viewing an area outside the vehicle, e.g. the exterior of the vehicle with a predetermined field of view to the sides of the vehicle
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/254—Analysis of motion involving subtraction of images
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G06T7/74—Determining position or orientation of objects or cameras using feature-based methods involving reference images or patches
    • 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
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
    • 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
    • G06V20/54—Surveillance or monitoring of activities, e.g. for recognising suspicious objects of traffic, e.g. cars on the road, trains or boats
    • 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
    • G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R2300/00—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle
    • B60R2300/80—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the intended use of the viewing arrangement
    • B60R2300/802—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the intended use of the viewing arrangement for monitoring and displaying vehicle exterior blind spot views
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
    • B60R2300/00—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle
    • B60R2300/80—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the intended use of the viewing arrangement
    • B60R2300/8093—Details of viewing arrangements using cameras and displays, specially adapted for use in a vehicle characterised by the intended use of the viewing arrangement for obstacle warning
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30248—Vehicle exterior or interior
    • G06T2207/30252—Vehicle exterior; Vicinity of vehicle
    • G06T2207/30261—Obstacle

Definitions

  • the present application relates to the field of vehicle warning technology, and in particular to a driving blind spot monitoring method, device, equipment and readable storage medium.
  • Blind spots refer to areas that the driver cannot directly see or observe while driving. These areas are located on the side and rear of the vehicle and are often prone to accidents. Blind spot monitoring is an important part of the assisted driving function. It can help the driver monitor obstacles in the blind spots, thereby reducing the potential dangers caused by them.
  • the current blind spot monitoring method uses a feature matching algorithm to identify obstacles from the video data collected by the camera.
  • the feature matching algorithm needs to be trained on a model to form a feature database before use. During use, it is limited by the feature database and can only identify a limited number of obstacles, which makes it easy for the blind spot monitoring method to miss detections.
  • the present application provides a driving blind spot monitoring method, device, equipment and readable storage medium, which can solve the technical problem that the driving blind spot monitoring method in the prior art is prone to missed detection.
  • an embodiment of the present application provides a driving blind spot monitoring method, the driving blind spot monitoring method comprising:
  • a grayscale change reference value of the target image group is less than a grayscale change threshold value
  • the grayscale change reference value of the target image group is calculated based on grayscale change quantization values of all two adjacent frames of images contained in the group
  • the grayscale change quantization value of two adjacent frames of images is determined based on inter-frame grayscale changes in a background area, and the greater the grayscale change, the greater the corresponding grayscale change quantization value
  • the grayscale average value of the same pixel in the target image group is used as the grayscale standard value of the pixel;
  • the obstacle recognition result in the obstacle monitoring area is output.
  • the step of determining the target image group from the target video data includes:
  • the grayscale change reference value of the N frames of image is calculated according to the first formula group, and the first formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • T f is the gray difference threshold of the same pixel in the background area of two adjacent frames
  • Ak (i, j) is the two-dimensional value of the gray difference between the pixel (i, j) in the k-th frame image and the k-1-th frame image
  • D is the background area
  • Ak is the quantized value of the gray change between the k-th frame image and the k-1-th frame image
  • A is the gray change reference value
  • the N frames of images are determined to be the target image group.
  • the obstacle monitoring area in the subsequent image of the target image group comprises:
  • the first difference reference value of the L frames of images is calculated according to the second formula group, and the second formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • R (i, j) is the gray standard value of the pixel (i, j)
  • T b is the gray difference threshold between the gray value of the same pixel in the obstacle monitoring area and the gray standard value
  • B k (i, j) is the gray difference two-dimensional value of the pixel (i, j) in the k-th frame image based on T b
  • d is the obstacle monitoring area
  • B k is the sum of the gray difference two-dimensional values of all pixels in the obstacle monitoring area of the k-th frame image
  • B is the first difference reference value
  • the first difference reference value of the L frames of images is less than the first difference threshold, it is determined that there is no obstacle in the obstacle monitoring area;
  • the first difference reference value of the L frames of images is greater than or equal to the first difference threshold, it is determined that an obstacle exists in the obstacle monitoring area.
  • the following steps are further included:
  • the average grayscale value of the same pixel in these N frames of images is used as the grayscale standard value of the pixel;
  • the method further includes:
  • the second difference reference value of the L frames of images is calculated according to the third formula group, and the third formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • R (i, j) is the gray standard value of the pixel (i, j)
  • TL is the gray difference threshold between the gray value of the same pixel in the background area and the gray standard value
  • C k (i, j) is the gray difference two-dimensional value of the pixel (i, j) in the k-th frame image based on TL
  • D is the background area
  • C k is the sum of the gray difference two-dimensional values of all pixels in the background area of the k-th frame image
  • C is the second difference reference value
  • the second difference reference value of the L frames of images is greater than or equal to the second difference threshold, then taking the first frame of the L frames of images as the starting frame, obtaining N consecutive frames of images from the target video data, and calculating the grayscale change reference value of the N frames of images, where N is equal to the number of image frames in the target image group;
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • the following step is further performed:
  • N is equal to the number of image frames in the target image group
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • the following step is further included:
  • the camera on the corresponding side of the vehicle is used as the target camera.
  • the background area includes all drivable areas in the area photographed by the target camera, and the obstacle monitoring area is located in the adjacent lane area corresponding to the target camera.
  • an embodiment of the present application further provides a driving blind spot monitoring device, the driving blind spot monitoring device comprising:
  • An area determination module is used to determine a background area and an obstacle monitoring area from the area photographed by the target camera, wherein the background area includes the obstacle monitoring area;
  • a preprocessing module is used to preprocess the video data collected by the target camera to obtain target video data
  • An image determination module is used to determine a target image group from target video data, wherein a grayscale change reference value of the target image group is less than a grayscale change threshold value, the grayscale change reference value of the target image group is calculated based on grayscale change quantization values of all two adjacent frames of images contained in the group, the grayscale change quantization value of two adjacent frames of images is determined based on inter-frame grayscale changes in a background area, and the greater the grayscale change, the greater the corresponding grayscale change quantization value;
  • a standard value calculation module is used to take the grayscale average value of each pixel in the obstacle monitoring area as the grayscale standard value of the pixel in the target image group;
  • the obstacle recognition module is used to output the obstacle recognition result in the obstacle monitoring area according to the difference between the grayscale value of all pixels in the obstacle monitoring area in the subsequent images of the target image group and the corresponding grayscale standard value.
  • the embodiment of the present application further provides a driving blind spot monitoring device, the driving blind spot monitoring device comprising a processor, a memory, and a program stored in the memory and executable by the processor.
  • a driving blind spot monitoring program wherein when the driving blind spot monitoring program is executed by the processor, the steps of the above-mentioned driving blind spot monitoring method are implemented.
  • an embodiment of the present application further provides a readable storage medium, on which a blind spot monitoring program is stored, wherein when the blind spot monitoring program is executed by a processor, the steps of the above-mentioned blind spot monitoring method are implemented.
  • This application obtains a target image group by analyzing the inter-frame grayscale changes in the background area of the target video data, calculates the grayscale standard value based on the target image group, and outputs the obstacle recognition result in the obstacle monitoring area according to the difference between the grayscale values of all pixels in the obstacle monitoring area in the subsequent images of the target image group and the corresponding grayscale standard value.
  • This application does not rely on feature matching and is not limited to the feature database, thereby reducing missed detections.
  • this application does not require model training before use, and has lower requirements for computing power and processing speed during use, which helps to reduce implementation costs.
  • FIG1 is a schematic diagram of a flow chart of a method for monitoring a vehicle blind spot in an embodiment of the present application
  • FIG2 is a schematic diagram of the process of obstacle recognition and road surface fluctuation recognition in one embodiment of the present application
  • FIG3 is a schematic diagram of a process of attempting to update a grayscale standard value in an embodiment of the present application
  • FIG4 is a schematic diagram of functional modules of a driving blind spot monitoring device according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of the hardware structure of the blind spot monitoring device involved in the embodiment of the present application.
  • A/B can mean A or B.
  • the “and/or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist.
  • a and/or B can mean: A exists alone, A and B exist at the same time, and B exists alone.
  • “multiple” refers to two or more than two.
  • an embodiment of the present application provides a method for monitoring a driving blind spot.
  • FIG. 1 is a schematic flow chart of a method for monitoring a vehicle's blind spot in an embodiment of the present application.
  • the driving blind spot monitoring method includes the following steps:
  • Blind spot detection systems usually monitor the area around the vehicle by using sensors, cameras or radars. When other vehicles or objects enter the blind spot, the system will sound an alarm to alert the driver.
  • the vision-based blind spot warning system currently on the market includes the following steps:
  • Camera angle adjustment Adjust the camera angle to cover the blind spot of the vehicle. Make sure the camera can clearly capture the situation on the side and rear.
  • Camera connection Connect the camera to the vehicle's electrical system.
  • the camera is usually connected to the display screen or navigation system on the vehicle.
  • Image processing and analysis The camera captures images around the vehicle in real time. These images are then processed by image processing and analysis algorithms to detect and identify obstacles in blind spots.
  • blind spot alarm Once the camera detects an obstacle in the blind spot, the system will trigger the alarm mechanism and remind the driver through the vehicle's display screen, sound or vibration.
  • a high-quality camera can provide clear and stable images and work properly under various lighting conditions.
  • the target camera in this embodiment belongs to the above-mentioned camera for blind spot monitoring. It can be understood that the area photographed by the target camera includes more than just the obstacle monitoring area that needs to be paid attention to during driving. Therefore, it is necessary to determine the obstacle monitoring area from the area photographed by the target camera, and its demarcation rules are determined according to vehicle characteristics and driving safety requirements.
  • the background area is an area including the obstacle monitoring area, which also includes other areas adjacent to the obstacle monitoring area, and is used to characterize the overall situation of the road surface where the obstacle monitoring area is located.
  • the target camera refers to the camera that needs to monitor video data.
  • the following step is further included:
  • the camera on the corresponding side of the vehicle is used as the target camera.
  • the target camera is determined according to the lane the vehicle is in. For example, if the vehicle is in the leftmost lane, there is no obstacle risk on its left side, and there is no need to monitor the video data of the left camera. There is an adjacent lane on its right side, and there is an obstacle risk, so it is necessary to monitor the video data of the right camera, and the right camera is used as the target camera.
  • the background area includes all visible Driving area
  • the obstacle monitoring area is located in the adjacent lane area corresponding to the target camera.
  • the background area is the entire road area (excluding trees, guardrails, etc.) in the area captured by the target camera
  • the obstacle monitoring area is a part of the adjacent lane area corresponding to the target camera.
  • the drivable area involved in the background area can be determined based on high-precision map data, or edge detection can be performed based on a visual segmentation algorithm.
  • the obstacle monitoring area is a rectangular area, the length of which is determined by vehicle characteristics and driving safety requirements, and the width of which is determined by the width of an adjacent lane.
  • the preprocessing process mainly includes time filtering and space filtering.
  • the purpose of time filtering is to obtain continuous video frames
  • the purpose of space filtering is to eliminate camera noise and environmental noise such as rain and snow.
  • the preprocessing process can also convert the video data into a form that is easy to process by subsequent algorithms, such as a gray matrix form.
  • determining a target image group from the target video data wherein a grayscale change reference value of the target image group is less than a grayscale change threshold value, the grayscale change reference value of the target image group is calculated based on grayscale change quantization values of all two adjacent frames of images contained in the group, the grayscale change quantization value of two adjacent frames of images is determined based on inter-frame grayscale changes in the background area, and the greater the grayscale change, the greater the corresponding grayscale change quantization value;
  • the grayscale change quantization value is determined according to the grayscale change between the background areas of two adjacent frames of images
  • the grayscale change reference value of a group of images is determined according to the grayscale change quantization values of all two adjacent frames of images of a group of continuous images.
  • the image group whose grayscale change reference value is less than the grayscale change threshold is taken as the target image group. Image group.
  • the grayscale change reference value of this embodiment is used to reflect the grayscale change of the background area of the continuous multi-frame images.
  • the smaller the grayscale change reference value the smaller the grayscale change and the lower the possibility of the existence of obstacles.
  • the larger the grayscale change reference value the larger the grayscale change and the higher the possibility of the existence of obstacles.
  • the grayscale change threshold is used to define whether there are obstacles in the background area.
  • the target image group is a group of continuous multi-frame images that are screened from the target video data based on the above judgment logic, which is considered that there are no obstacles in the background area.
  • the step of determining the target image group from the target video data includes:
  • the grayscale change reference value of the N frames of image is calculated according to the first formula group, and the first formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • T f is the gray difference threshold of the same pixel in the background area of two adjacent frames
  • Ak (i, j) is the two-dimensional value of the gray difference between the pixel (i, j) in the k-th frame image and the k-1-th frame image
  • D is the background area
  • Ak is the quantized value of the gray change between the k-th frame image and the k-1-th frame image
  • A is the gray change reference value
  • the N frames of images are determined to be the target image group.
  • the grayscale change reference value of the N frames of images is greater than or equal to the grayscale change threshold, the first frame of the N frames of images is discarded, and the next frame of the Nth frame of images is acquired, and the grayscale change reference value of the new N frames of images is calculated until the target image group is successfully acquired. That is, a sliding window with a sliding step of 1 frame and a size of N frames is created to continuously acquire N consecutive frames of images until the target image group is successfully acquired.
  • the grayscale standard value is used to characterize the grayscale value when there is no obstacle in the obstacle monitoring area.
  • the appropriate expansion of the background area relative to the obstacle monitoring area helps to improve the reliability of the grayscale standard value. Assuming that the target image group is determined and the grayscale standard value is calculated directly based on the grayscale change of the obstacle monitoring area, it is possible that an obstacle will be identified as a stable road surface.
  • the grayscale standard value is calculated according to the following formula:
  • R(i,j) is the grayscale standard value of pixel (i,j)
  • f k (i,j) is the grayscale value of pixel (i,j) in the k-th frame image.
  • the grayscale standard value cannot be calculated, and the subsequent obstacle recognition operation cannot be performed.
  • obstacle recognition can be performed with the help of other detection methods, such as radar detection, or using the preset grayscale standard value.
  • the grayscale values of all pixels in the obstacle monitoring area in the subsequent image are The greater the difference in the grayscale standard values, the higher the possibility that an obstacle exists, and the smaller the difference, the lower the possibility that an obstacle exists.
  • the step of outputting the obstacle recognition result in the obstacle monitoring area according to the difference between the grayscale values of all pixels in the obstacle monitoring area in the subsequent images of the target image group and the corresponding grayscale standard values comprises:
  • the first difference reference value of the L frames of images is calculated according to the second formula group, and the second formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • R (i, j) is the gray standard value of the pixel (i, j)
  • T b is the gray difference threshold between the gray value of the same pixel in the obstacle monitoring area and the gray standard value
  • B k (i, j) is the gray difference two-dimensional value of the pixel (i, j) in the k-th frame image based on T b
  • d is the obstacle monitoring area
  • B k is the sum of the gray difference two-dimensional values of all pixels in the obstacle monitoring area of the k-th frame image
  • B is the first difference reference value
  • the first difference reference value of the L frames of images is less than the first difference threshold, it is determined that there is no obstacle in the obstacle monitoring area;
  • the first difference reference value of the L frames of images is greater than or equal to the first difference threshold, it is determined that an obstacle exists in the obstacle monitoring area.
  • the first difference reference value is used to reflect the overall average difference between the grayscale of all pixels in the obstacle monitoring area of L consecutive frames of images and the corresponding grayscale standard value.
  • the first difference threshold is used to determine whether there is an obstacle in the background area. Compared with the method of monitoring each image individually, the missed detection rate is lower.
  • this embodiment obtains a target image group by analyzing the inter-frame grayscale changes of the background area in the target video data, calculates the grayscale standard value based on the target image group, and outputs the obstacle recognition result in the obstacle monitoring area according to the difference between the grayscale values of all pixels in the obstacle monitoring area in the subsequent images of the target image group and the corresponding grayscale standard value.
  • This embodiment does not rely on feature matching and is not limited to the feature database, thereby reducing missed detection.
  • this embodiment does not require model training before use, and has lower requirements for computing power and processing speed during use, which helps to reduce implementation costs.
  • the accuracy of the obstacle recognition algorithm in this application depends on the stability of the road surface appearance. For scenes where the road surface appearance changes frequently, its accuracy is difficult to guarantee.
  • the highway area is a typical area with a stable road surface appearance. The application works well in the highway area.
  • FIG2 is a schematic diagram showing the process of obstacle recognition and road surface fluctuation recognition in one embodiment of the present application
  • FIG3 is a schematic diagram showing the process of attempting to update the grayscale standard value in one embodiment of the present application.
  • the following steps are further included:
  • the average grayscale value of the same pixel in these N frames of images is used as the grayscale standard value of the pixel;
  • the method further includes:
  • the second difference reference value of the L frames of images is calculated according to the third formula group, and the third formula group is:
  • f k (i, j) is the gray value of pixel (i, j) in the k-th frame image
  • R (i, j) is the gray value of pixel (i, j) Grayscale standard value
  • TL is the grayscale difference threshold between the grayscale value of the same pixel in the background area and the grayscale standard value
  • Ck (i,j) is the grayscale difference two-dimensional value obtained based on TL for the pixel ( i,j) in the k-th frame image
  • D is the background area
  • Ck is the sum of the grayscale difference two-dimensional values of all pixels in the background area of the k-th frame image
  • C is the second difference reference value
  • the second difference reference value of the L frames of images is greater than or equal to the second difference threshold, then taking the first frame of the L frames of images as the starting frame, obtaining N consecutive frames of images from the target video data, and calculating the grayscale change reference value of the N frames of images, where N is equal to the number of image frames in the target image group;
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • This embodiment provides a method for updating the target image group and grayscale standard value.
  • the second difference reference value of the L-frame image is further calculated.
  • the second difference reference value is used to reflect the overall average difference between the grayscale of all pixels in the background area of the continuous L-frame image and the corresponding grayscale standard value.
  • the second difference threshold is used to define whether there is road surface fluctuation in the background area. For example, due to factors such as road surface material or external lighting, the overall color of the road surface in the image changes, and there may also be obstacles in the area outside the obstacle monitoring area.
  • the target image group and the grayscale standard value will be tried to be updated.
  • the following step is further performed:
  • N is equal to the number of image frames in the target image group
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • This embodiment provides another method for updating the target image group and the grayscale standard value.
  • a trial operation is performed, that is, N consecutive frames of images are acquired, and the grayscale change reference value is calculated. If the grayscale change reference value meets the requirement, the target image group and the grayscale standard value are updated, otherwise no update is performed. For example, an update attempt is performed every 30 minutes of driving of the vehicle.
  • the first real-time fluctuation update strategy ensures that the grayscale standard value is updated in time when the road surface changes to a certain extent.
  • the second periodic update strategy ensures that a certain update frequency can be guaranteed even when the road surface changes slightly, thereby comprehensively improving the reliability of the grayscale standard value and ensuring the accuracy of the obstacle recognition algorithm.
  • an embodiment of the present application also provides a driving blind spot monitoring device.
  • FIG. 4 is a schematic diagram showing the functional modules of a driving blind spot monitoring device in an embodiment of the present application.
  • the driving blind spot monitoring device includes:
  • the area determination module 10 is used to determine the background area and the obstacle monitoring area from the area photographed by the target camera, wherein the background area includes the obstacle monitoring area;
  • the preprocessing module 20 is used to preprocess the video data collected by the target camera to obtain target video data;
  • the image determination module 30 is used to determine the target image group from the target video data, wherein the grayscale change reference value of the target image group is less than the grayscale change threshold value, the grayscale change reference value of the target image group is calculated based on the grayscale change quantization value of all two adjacent frames of images contained in the group, and the grayscale change quantization value of the two adjacent frames of images is determined based on the inter-frame grayscale change of the background area.
  • the larger the grayscale change the corresponding The larger the grayscale change quantization value is;
  • the standard value calculation module 40 is used to use the grayscale average value of each pixel in the obstacle monitoring area as the grayscale standard value of the pixel in the target image group;
  • the obstacle recognition module 50 is used to output the obstacle recognition result in the obstacle monitoring area according to the difference between the grayscale values of all pixels in the obstacle monitoring area in the subsequent images of the target image group and the corresponding grayscale standard values.
  • the image determination module 30 is used to:
  • the grayscale change reference value of the N frames of image is calculated according to the first formula group, and the first formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • T f is the gray difference threshold of the same pixel in the background area of two adjacent frames
  • Ak (i, j) is the two-dimensional value of the gray difference between the pixel (i, j) in the k-th frame image and the k-1-th frame image
  • D is the background area
  • Ak is the quantized value of the gray change between the k-th frame image and the k-1-th frame image
  • A is the gray change reference value
  • the N frames of images are determined to be the target image group.
  • the obstacle identification module 50 is used to:
  • the first difference reference value of the L frames of images is calculated according to the second formula group, and the second formula group is:
  • f k (i, j) is the gray value of the pixel (i, j) in the k-th frame image
  • R (i, j) is the gray standard value of the pixel (i, j)
  • T b is the gray difference threshold between the gray value of the same pixel in the obstacle monitoring area and the gray standard value
  • B k (i, j) is the gray difference two-dimensional value of the pixel (i, j) in the k-th frame image based on T b
  • d is the obstacle monitoring area
  • B k is the sum of the gray difference two-dimensional values of all pixels in the obstacle monitoring area of the k-th frame image
  • B is the first difference reference value
  • the first difference reference value of the L frames of images is less than the first difference threshold, it is determined that there is no obstacle in the obstacle monitoring area;
  • the first difference reference value of the L frames of images is greater than or equal to the first difference threshold, it is determined that an obstacle exists in the obstacle monitoring area.
  • the standard value calculation module 40 is also used for:
  • the average grayscale value of the same pixel in these N frames of images is used as the grayscale standard value of the pixel;
  • the driving blind spot monitoring device also includes an update module for:
  • the second difference reference value of the L frames of images is calculated according to the third formula group, and the third formula group is:
  • f k (u, j) is the gray value of the pixel (i, j) in the k-th frame image
  • R (i, j) is the gray standard value of the pixel (i, j)
  • TL is the gray difference threshold between the gray value of the same pixel in the background area and the gray standard value
  • C k (i, j) is the gray difference two-dimensional value of the pixel (i, j) in the k-th frame image based on TL
  • D is the background area
  • C k is the sum of the gray difference two-dimensional values of all pixels in the background area of the k-th frame image
  • C is the second difference reference value
  • the second difference reference value of the L frames of images is greater than or equal to the second difference threshold, then taking the first frame of the L frames of images as the starting frame, obtaining N consecutive frames of images from the target video data, and calculating the grayscale change reference value of the N frames of images, where N is equal to the number of image frames in the target image group;
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • the driving blind spot monitoring device further includes an updating module, which is used to:
  • N is equal to the number of image frames in the target image group
  • the grayscale change reference value of the N frames of images is less than the grayscale change threshold, the N frames of images are used as a new target image group, and the grayscale standard value is updated based on the new target image group.
  • the driving blind spot monitoring device further includes a camera determination module, which is used to:
  • the camera on the corresponding side of the vehicle is used as the target camera.
  • the background area includes all drivable areas in the area photographed by the target camera, and the obstacle monitoring area is located in the adjacent lane area corresponding to the target camera.
  • each module in the above-mentioned blind spot monitoring device corresponds to the various steps in the above-mentioned blind spot monitoring method embodiment, and its functions and implementation processes will not be repeated here one by one.
  • an embodiment of the present application provides a blind spot monitoring device, which can be a personal computer (PC), a laptop computer, a server, or other device with data processing capabilities.
  • a blind spot monitoring device which can be a personal computer (PC), a laptop computer, a server, or other device with data processing capabilities.
  • FIG5 shows a schematic diagram of the hardware structure of the blind spot monitoring device involved in the embodiment of the present application.
  • the blind spot monitoring device may include a processor, a memory, a communication interface, and a communication bus.
  • the communication bus may be of any type and is used to interconnect the processor, the memory, and the communication interface.
  • the communication interface includes input/output (I/O) interface, physical interface and logical interface, etc., which are used to realize the interconnection of devices inside the blind spot monitoring device, and the interface used to realize the interconnection between the blind spot monitoring device and other devices (such as other computing devices or user devices).
  • the physical interface can be an Ethernet interface, a fiber optic interface, an ATM interface, etc.; the user device can be a display, a keyboard, etc.
  • the memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.
  • RAM random access memory
  • ROM read-only memory
  • NVRAM non-volatile RAM
  • flash memory optical storage
  • hard disk programmable ROM
  • PROM erasable PROM
  • EEPROM electrically erasable PROM
  • the processor may be a general-purpose processor, which may call the blind spot monitoring program stored in the memory and execute the blind spot monitoring method provided in the embodiment of the present application.
  • the general-purpose processor may be a central processing unit (CPU).
  • the method executed when the blind spot monitoring program is called may refer to the various embodiments of the blind spot monitoring method of the present application, and will not be described here. Elaborate.
  • FIG. 5 does not constitute a limitation on the present application, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.
  • an embodiment of the present application also provides a readable storage medium.
  • the readable storage medium of the present application stores a driving blind spot monitoring program, wherein when the driving blind spot monitoring program is executed by the processor, the steps of the driving blind spot monitoring method as described above are implemented.
  • the method implemented when the blind spot monitoring program is executed can refer to the various embodiments of the blind spot monitoring method of the present application, and will not be repeated here.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • Databases & Information Systems (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Mechanical Engineering (AREA)
  • Closed-Circuit Television Systems (AREA)

Abstract

本申请提供一种行车盲区监测方法、装置、设备及可读存储介质，该方法包括：从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域；对目标摄像头采集的视频数据进行预处理，得到目标视频数据；从目标视频数据中确定目标图像组；针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。本申请不依赖特征匹配，不会受限于特征数据库，从而减少漏检情况。

Description

行车盲区监测方法、装置、设备及可读存储介质 技术领域
本申请涉及车辆预警技术领域，具体涉及一种行车盲区监测方法、装置、设备及可读存储介质。
背景技术
行车盲区是指驾驶员在驾驶过程中无法直接看到或观察到的区域，这些区域位于车辆的侧后方，往往容易发生事故。行车盲区监测是辅助驾驶功能中的一个重要部分，它可以帮助驾驶员监测行车盲区中的障碍物情况，从而减少其造成的潜在危险。目前的行车盲区监测方法采用特征匹配算法从摄像头采集的视频数据中识别出障碍物，特征匹配算法在使用前需要进行模型训练形成特征数据库，在使用过程中受限于特征数据库，可识别障碍物种类有限，导致行车盲区监测方法容易出现漏检情况。
发明内容
本申请提供一种行车盲区监测方法、装置、设备及可读存储介质，可以解决现有技术中的行车盲区监测方法容易出现漏检情况的技术问题。
第一方面，本申请实施例提供一种行车盲区监测方法，所述行车盲区监测方法包括：
从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的灰度变化量化值越大；
针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
进一步地，一实施例中，所述从目标视频数据中确定目标图像组的步骤包括：
从目标视频数据中获取连续N帧图像；
根据第一公式组计算得到这N帧图像的灰度变化参考值，第一公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，Tf为相邻两帧图像的背景区域中同一像素点的灰度差异阈值，Ak(i,j)为像素点(i,j)在第k帧图像与第k-1帧图像之间的灰度差异二维值，D为背景区域，Ak为第k帧图像与第k-1帧图像的灰度变化量化值，A为灰度变化参考值；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则确定这N帧图像为目标图像组。
进一步地，一实施例中，所述根据目标图像组后续的图像中障碍物监测区 域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果的步骤包括：
从目标图像组后续的图像中获取连续L帧图像；
根据第二公式组计算得到这L帧图像的第一差异参考值，第二公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，Tb为障碍物监测区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Bk(i,j)为第k帧图像中的像素点(i,j)基于Tb得到的灰度差异二维值，d为障碍物监测区域，Bk为第k帧图像的障碍物监测区域中所有像素点的灰度差异二维值之和，B为第一差异参考值；
若这L帧图像的第一差异参考值小于第一差异阈值，则判断障碍物监测区域不存在障碍物；
若这L帧图像的第一差异参考值大于或等于第一差异阈值，则判断障碍物监测区域存在障碍物。
进一步地，一实施例中，在所述从目标视频数据中确定目标图像组的步骤之后还包括：
针对背景区域中除障碍物监测区域以外的其他每个像素点，将同一像素点在这N帧图像中的灰度平均值作为该像素点的灰度标准值；
在所述判断障碍物监测区域不存在障碍物的步骤之后还包括：
根据第三公式组计算得到这L帧图像的第二差异参考值，第三公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，TL为背景区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Ck(i,j)为第k帧图像中的像素点(i,j)基于TL得到的灰度差异二维值，D为背景区域，Ck为第k帧图像的背景区域中所有像素点的灰度差异二维值之和，C为第二差异参考值；
若这L帧图像的第二差异参考值大于或等于第二差异阈值，则以这L帧图像的第一帧为起始帧，从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
进一步地，一实施例中，在所述针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值的步骤之后还包括：
根据更新周期定期从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
进一步地，一实施例中，在所述从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域的步骤之前还包括：
根据车辆定位系统和高精度地图确定车辆所在车道；
若车辆所在车道的左侧和/或右侧存在相邻车道，则将车辆对应侧的摄像头作为目标摄像头。
进一步地，一实施例中，背景区域包含目标摄像头拍摄的区域中的所有可行驶区域，障碍物监测区域位于目标摄像头对应的相邻车道区域中。
第二方面，本申请实施例还提供一种行车盲区监测装置，所述行车盲区监测装置包括：
区域确定模块，用于从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
预处理模块，用于对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
图像确定模块，用于从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的灰度变化量化值越大；
标准值计算模块，用于针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
障碍物识别模块，用于根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
第三方面，本申请实施例还提供一种行车盲区监测设备，所述行车盲区监测设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行 的行车盲区监测程序，其中所述行车盲区监测程序被所述处理器执行时，实现上述行车盲区监测方法的步骤。
第四方面，本申请实施例还提供一种可读存储介质，所述可读存储介质上存储有行车盲区监测程序，其中所述行车盲区监测程序被处理器执行时，实现上述行车盲区监测方法的步骤。
本申请通过分析目标视频数据中背景区域的帧间灰度变化情况得到目标图像组，基于目标图像组计算得到灰度标准值，根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。本申请不依赖特征匹配，不会受限于特征数据库，从而减少漏检情况。另外，本申请不需要在使用前进行模型训练，且使用过程中对于计算能力与处理速度的要求更低，有助于降低实施成本。
附图说明
图1为本申请一实施例中行车盲区监测方法的流程示意图；
图2为本申请一实施例中障碍物识别和路面波动识别的流程示意图；
图3为本申请一实施例中尝试更新灰度标准值的流程示意图；
图4为本申请一实施例中行车盲区监测装置的功能模块示意图；
图5为本申请实施例方案中涉及的行车盲区监测设备的硬件结构示意图。
具体实施方式
为了使本技术领域的人员更好地理解本申请方案，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请 中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
本申请的说明书和权利要求书及上述附图中的术语“包括”和“具有”以及它们任何变形，意图在于覆盖不排他的包含。例如包含了一系列步骤或单元的过程、方法、系统、产品或设备没有限定于已列出的步骤或单元，而是可选地还包括没有列出的步骤或单元，或可选地还包括对于这些过程、方法、产品或设备固有的其他步骤或单元。术语“第一”、“第二”和“第三”等描述，是用于区分不同的对象等，其不代表先后顺序，也不限定“第一”、“第二”和“第三”是不同的类型。
在本申请实施例的描述中，“示例性的”、“例如”或者“举例来说”等用于表示作例子、例证或说明。本申请实施例中被描述为“示例性的”、“例如”或者“举例来说”的任何实施例或设计方案不应被解释为比其它实施例或设计方案更优选或更具优势。确切而言，使用“示例性的”、“例如”或者“举例来说”等词旨在以具体方式呈现相关概念。
在本申请实施例的描述中，除非另有说明，“/”表示或的意思，例如，A/B可以表示A或B；文本中的“和/或”仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况，另外，在本申请实施例的描述中，“多个”是指两个或多于两个。
在本申请实施例描述的一些流程中，包含了按照特定顺序出现的多个操作或步骤，但是应该理解，这些操作或步骤可以不按照其在本申请实施例中出现的顺序来执行或并行执行，操作的序号仅用于区分开各个不同的操作，序号本身不代表任何的执行顺序。另外，这些流程可以包括更多或更少的操作，并且 这些操作或步骤可以按顺序执行或并行执行，并且这些操作或步骤可以进行组合。
为使本申请的目的、技术方案和优点更加清楚，下面将结合附图对本申请实施方式作进一步地详细描述。
第一方面，本申请实施例提供一种行车盲区监测方法。
图1示出了本申请一实施例中行车盲区监测方法的流程示意图。
参照图1，一实施例中，行车盲区监测方法包括如下步骤：
S11、从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
行车盲区检测系统通常通过使用传感器、摄像头或雷达来监测车辆周围的区域。当有其他车辆或物体进入盲区时，系统会发出警报，提醒驾驶员注意。目前市场上基于视觉的行车盲区预警系统包括以下步骤：
1、安装摄像头：首先，需要在车辆的侧后方适当位置安装摄像头。一般来说，摄像头会安装在车辆的后视镜下方或车身侧面。
2、摄像头视角调整：调整摄像头的视角，使其能够覆盖到车辆的盲区区域。确保摄像头能够清晰地拍摄到侧后方的情况。
3、摄像头连接：将摄像头与车辆的电气系统连接起来。摄像头通常会与车辆上的显示屏或导航系统相连。
4、图像处理和分析：摄像头会实时捕捉车辆周围的图像。然后，这些图像会经过图像处理和分析算法的处理，用于检测和识别盲区中的障碍物。
5、盲区警报：一旦摄像头检测到盲区有障碍物存在，系统会触发警报机制，通过车辆上的显示屏、声音或震动等方式提醒驾驶员。
6、可视化显示：一些车辆上的显示屏会显示摄像头捕捉到的图像，以帮助 驾驶员更直观地了解盲区的情况。
需要注意的是，摄像头的性能和质量对盲区监测的准确性和可靠性有很大影响。高质量的摄像头能够提供清晰、稳定的图像，并能在各种光照条件下正常工作。
本实施例中的目标摄像头属于上述用于盲区监测的摄像头。可以理解，目标摄像头拍摄的区域不止包含行车过程中需要关注的障碍物监测区域，因此需要从目标摄像头拍摄的区域中确定障碍物监测区域，其划定规则根据车辆特性和行车安全需求确定。背景区域是包含障碍物监测区域在内的区域，其还包含与障碍物监测区域邻接的其他区域，用于表征障碍物监测区域所在路面的整体情况。
特别说明，车辆行驶过程中并不需要全程监测车辆左侧和右侧两个摄像头的视频数据，例如，车辆其中一侧不存在障碍物风险，则无需监测该侧摄像头的视频数据。本实施例中目标摄像头是指需要进行视频数据监测的摄像头。
进一步地，一实施例中，在所述从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域的步骤之前还包括：
根据车辆定位系统和高精度地图确定车辆所在车道；
若车辆所在车道的左侧和/或右侧存在相邻车道，则将车辆对应侧的摄像头作为目标摄像头。
本实施例中，根据车辆所在车道情况确定目标摄像头。例如，车辆处于最左侧的车道，其左侧不存在障碍物风险，无需监测左侧摄像头的视频数据，其右侧存在相邻车道，存在障碍物风险，需要监测右侧摄像头的视频数据，右侧摄像头作为目标摄像头。
进一步地，一实施例中，背景区域包含目标摄像头拍摄的区域中的所有可 行驶区域，障碍物监测区域位于目标摄像头对应的相邻车道区域中。
本实施例中，背景区域也即目标摄像头拍摄的区域中的整个道路区域(不包含树木、护栏等)，障碍物监测区域是目标摄像头对应的相邻车道区域的一部分。例如，车辆的左侧存在三个车道，均包含于目标摄像头的拍摄区域内，则这三个车道区域都包含于背景区域，仅左侧相邻车道的部分区域作为障碍物监测区域。
可选地，背景区域涉及的可行驶区域可根据高精度地图数据确定，也可基于视觉分割算法进行边缘检测。
可选地，障碍物监测区域为矩形区域，其长度由车辆特性和行车安全需求确定，其宽度由相邻车道宽度确定。
S12、对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
本实施例中，预处理过程主要包括时间滤波和空间滤波，时间滤波的目的是得到连续的视频帧，空间滤波的目的是消除摄像机噪声以及雨雪等环境噪声。可选地，预处理过程还可将视频数据转换为后续算法易处理的形式，例如灰度矩阵形式。
S13、从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的灰度变化量化值越大；
本实施例中，根据相邻两帧图像背景区域的帧间灰度变化情况确定灰度变化量化值，以一组连续图像的所有相邻两帧图像的灰度变化量化值确定这组图像的灰度变化参考值，将灰度变化参考值小于灰度变化阈值的图像组作为目标 图像组。
可以理解，假设一段时间内摄像头拍摄的连续多帧图像的背景区域中不存在障碍物、只包含路面，且这段时间的路面外观基本稳定不变，那么这些图像中背景区域的灰度也基本稳定不变，本实施例的灰度变化参考值即用于反映连续多帧图像背景区域的灰度变化情况，灰度变化参考值越小，代表灰度变化越小，存在障碍物的可能性越低，灰度变化参考值越大，代表灰度变化越大，存在障碍物的可能性越高。灰度变化阈值用于界定背景区域中是否存在障碍物。目标图像组是基于上述判断逻辑，从目标视频数据中筛选出的一组认为背景区域中不存在障碍物的连续多帧图像。
进一步地，一实施例中，所述从目标视频数据中确定目标图像组的步骤包括：
从目标视频数据中获取连续N帧图像；
根据第一公式组计算得到这N帧图像的灰度变化参考值，第一公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，Tf为相邻两帧图像的背景区域中同一像素点的灰度差异阈值，Ak(i,j)为像素点(i,j)在第k帧图像与第k-1帧图像之间的灰度差异二维值，D为背景区域，Ak为第k帧图像与第k-1帧图像的灰度变化量化值，A为灰度变化参考值；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则确定这N帧图像为目标图像组。
本实施例中，对于车辆启动初期，还没有成功获取到目标图像组的情况下，若这N帧图像的灰度变化参考值大于或等于灰度变化阈值，则丢弃这N帧图像的第一帧，再获取第N帧图像的后一帧，计算新的N帧图像的灰度变化参考值，直到成功获取到目标图像组。即，创建滑动步长为1帧、大小为N帧的滑动窗口不断获取连续N帧图像，直到成功获取到目标图像组。
S14、针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
本实施例中，灰度标准值用于表征障碍物监测区域不存在障碍物时的灰度值。背景区域相对于障碍物监测区域的适当外扩，有助于提高灰度标准值的可靠性。假设直接根据障碍物监测区域的灰度变化情况确定目标图像组并计算灰度标准值，有可能会将某个障碍物识别为稳定路面。
具体地，根据下列公式计算得到灰度标准值：
其中，R(i,j)为像素点(i,j)的灰度标准值，fk(i,j)为第k帧图像中像素点(i,j)的灰度值。
对于车辆启动初期，还没有成功获取到目标图像组的情况下，无法计算出灰度标准值，也就无法执行后续障碍物识别操作。此时可借助其他检测手段进行障碍物识别，例如雷达检测，或者采用预置的灰度标准值。在已经成功获取到目标图像组并计算灰度标准值后，还会涉及到两者的更新，该部分将在后文中详细展开。
S15、根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
本实施例中，后续图像中障碍物监测区域的所有像素点的灰度值与对应的 灰度标准值的差异越大，存在障碍物的可能性越高，差异越小，存在障碍物的可能性越低。
进一步地，一实施例中，所述根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果的步骤包括：
从目标图像组后续的图像中获取连续L帧图像；
根据第二公式组计算得到这L帧图像的第一差异参考值，第二公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，Tb为障碍物监测区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Bk(i,j)为第k帧图像中的像素点(i,j)基于Tb得到的灰度差异二维值，d为障碍物监测区域，Bk为第k帧图像的障碍物监测区域中所有像素点的灰度差异二维值之和，B为第一差异参考值；
若这L帧图像的第一差异参考值小于第一差异阈值，则判断障碍物监测区域不存在障碍物；
若这L帧图像的第一差异参考值大于或等于第一差异阈值，则判断障碍物监测区域存在障碍物。
本实施例中，第一差异参考值用于反映连续L帧图像的障碍物监测区域的所有像素点的灰度与对应的灰度标准值的整体平均差异情况，第一差异阈值用于界定背景区域中是否存在障碍物，相比逐张图像单独监测的方式漏检率更低。
由此，本实施例通过分析目标视频数据中背景区域的帧间灰度变化情况得到目标图像组，基于目标图像组计算得到灰度标准值，根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。本实施例不依赖特征匹配，不会受限于特征数据库，从而减少漏检情况。另外，本实施例不需要在使用前进行模型训练，且使用过程中对于计算能力与处理速度的要求更低，有助于降低实施成本。
需要说明的是，本申请中障碍物识别算法的准确性依赖于路面外观的稳定，对于路面外观频繁变化的场景，其准确性难以保证。高速公路区域是典型的路面外观稳定的区域，本申请在高速公路区域的使用效果良好。
图2示出了本申请一实施例中障碍物识别和路面波动识别的流程示意图；图3示出了本申请一实施例中尝试更新灰度标准值的流程示意图。
进一步地，一实施例中，在所述从目标视频数据中确定目标图像组的步骤之后还包括：
针对背景区域中除障碍物监测区域以外的其他每个像素点，将同一像素点在这N帧图像中的灰度平均值作为该像素点的灰度标准值；
在所述判断障碍物监测区域不存在障碍物的步骤之后还包括：
根据第三公式组计算得到这L帧图像的第二差异参考值，第三公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的 灰度标准值，TL为背景区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Ck(i,j)为第k帧图像中的像素点(i,j)基于TL得到的灰度差异二维值，D为背景区域，Ck为第k帧图像的背景区域中所有像素点的灰度差异二维值之和，C为第二差异参考值；
若这L帧图像的第二差异参考值大于或等于第二差异阈值，则以这L帧图像的第一帧为起始帧，从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
本实施例提供一种更新目标图像组和灰度标准值的方法，参照图2和图3，在计算L帧图像的第一差异参考值，判断不存在障碍物后，还会进一步计算这L帧图像的第二差异参考值，第二差异参考值用于反映连续L帧图像的背景区域的所有像素点的灰度与对应的灰度标准值的整体平均差异情况，第二差异阈值用于界定背景区域中是否存在路面波动，例如，受路面材质或外部光照等因素影响，图像中路面整体颜色发生改变，也有可能是障碍物监测区域以外的区域存在障碍物。根据第二差异参考值判断存在路面波动时，则会尝试更新目标图像组和灰度标准值。
需要说明的是，判断存在路面波动时，仅执行一次尝试操作，即获取连续N帧图像，计算灰度变化参考值，若灰度变化参考值满足要求，则更新目标图像组和灰度标准值，否则不进行更新。
进一步地，一实施例中，在所述针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值的步骤之后还包括：
根据更新周期定期从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
本实施例提供另一种更新目标图像组和灰度标准值的方法，参照图3，在达到更新周期时，执行一次尝试操作，即获取连续N帧图像，计算灰度变化参考值，若灰度变化参考值满足要求，则更新目标图像组和灰度标准值，否则不进行更新。例如，车辆每行驶30分钟进行一次更新尝试。
需要说明的是，上述两种更新方法可结合使用，第一种实时波动更新策略保证路面变化到一定程度时及时更新灰度标准值，第二种定期更新策略保证路面变化较小时也能够进行保证一定更新频率，从而综合提高灰度标准值的可靠性，保证障碍物识别算法的准确性。
第二方面，本申请实施例还提供一种行车盲区监测装置。
图4示出了本申请一实施例中行车盲区监测装置的功能模块示意图。
参照图4，一实施例中，行车盲区监测装置包括：
区域确定模块10，用于从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
预处理模块20，用于对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
图像确定模块30，用于从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的 灰度变化量化值越大；
标准值计算模块40，用于针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
障碍物识别模块50，用于根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
进一步地，一实施例中，图像确定模块30用于：
从目标视频数据中获取连续N帧图像；
根据第一公式组计算得到这N帧图像的灰度变化参考值，第一公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，Tf为相邻两帧图像的背景区域中同一像素点的灰度差异阈值，Ak(i,j)为像素点(i,j)在第k帧图像与第k-1帧图像之间的灰度差异二维值，D为背景区域，Ak为第k帧图像与第k-1帧图像的灰度变化量化值，A为灰度变化参考值；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则确定这N帧图像为目标图像组。
进一步地，一实施例中，障碍物识别模块50用于：
从目标图像组后续的图像中获取连续L帧图像；
根据第二公式组计算得到这L帧图像的第一差异参考值，第二公式组为：


其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，Tb为障碍物监测区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Bk(i,j)为第k帧图像中的像素点(i,j)基于Tb得到的灰度差异二维值，d为障碍物监测区域，Bk为第k帧图像的障碍物监测区域中所有像素点的灰度差异二维值之和，B为第一差异参考值；
若这L帧图像的第一差异参考值小于第一差异阈值，则判断障碍物监测区域不存在障碍物；
若这L帧图像的第一差异参考值大于或等于第一差异阈值，则判断障碍物监测区域存在障碍物。
进一步地，一实施例中，标准值计算模块40还用于：
针对背景区域中除障碍物监测区域以外的其他每个像素点，将同一像素点在这N帧图像中的灰度平均值作为该像素点的灰度标准值；
行车盲区监测装置还包括更新模块，用于：
根据第三公式组计算得到这L帧图像的第二差异参考值，第三公式组为：


其中，fk(u,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，TL为背景区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Ck(i,j)为第k帧图像中的像素点(i,j)基于TL得到的灰度差异二维值，D为背景区域，Ck为第k帧图像的背景区域中所有像素点的灰度差异二维值之和，C为第二差异参考值；
若这L帧图像的第二差异参考值大于或等于第二差异阈值，则以这L帧图像的第一帧为起始帧，从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
进一步地，一实施例中，行车盲区监测装置还包括更新模块，用于：
根据更新周期定期从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
进一步地，一实施例中，行车盲区监测装置还包括摄像头确定模块，用于：
根据车辆定位系统和高精度地图确定车辆所在车道；
若车辆所在车道的左侧和/或右侧存在相邻车道，则将车辆对应侧的摄像头作为目标摄像头。
进一步地，一实施例中，背景区域包含目标摄像头拍摄的区域中的所有可行驶区域，障碍物监测区域位于目标摄像头对应的相邻车道区域中。
其中，上述行车盲区监测装置中各个模块的功能实现与上述行车盲区监测方法实施例中各步骤相对应，其功能和实现过程在此处不再一一赘述。
第三方面，本申请实施例提供一种行车盲区监测设备，行车盲区监测设备可以是个人计算机(personal computer，PC)、笔记本电脑、服务器等具有数据处理功能的设备。
图5示出了本申请实施例方案中涉及的行车盲区监测设备的硬件结构示意图。
参照图5，本申请实施例中，行车盲区监测设备可以包括处理器、存储器、通信接口以及通信总线。
其中，通信总线可以是任何类型的，用于实现处理器、存储器以及通信接口互连。
通信接口包括输入/输出(input/output，I/O)接口、物理接口和逻辑接口等用于实现行车盲区监测设备内部的器件互连的接口，以及用于实现行车盲区监测设备与其他设备(例如其他计算设备或用户设备)互连的接口。物理接口可以是以太网接口、光纤接口、ATM接口等；用户设备可以是显示屏(Display)、键盘(Keyboard)等。
存储器可以是各种类型的存储介质，例如随机存取存储器(random access memory，RAM)、只读存储器(read-only memory，ROM)、非易失性RAM(non-volatile RAM，NVRAM)、闪存、光存储器、硬盘、可编程ROM(programmable ROM，PROM)、可擦除PROM(erasable PROM，EPROM)、电可擦除PROM(electrically erasable PROM，EEPROM)等。
处理器可以是通用处理器，通用处理器可以调用存储器中存储的行车盲区监测程序，并执行本申请实施例提供的行车盲区监测方法。例如，通用处理器可以是中央处理器(central processing unit，CPU)。其中，行车盲区监测程序被调用时所执行的方法可参照本申请行车盲区监测方法的各个实施例，此处不再 赘述。
本领域技术人员可以理解，图5中示出的硬件结构并不构成对本申请的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
第四方面，本申请实施例还提供一种可读存储介质。
本申请可读存储介质上存储有行车盲区监测程序，其中所述行车盲区监测程序被处理器执行时，实现如上述的行车盲区监测方法的步骤。
其中，行车盲区监测程序被执行时所实现的方法可参照本申请行车盲区监测方法的各个实施例，此处不再赘述。
需要说明的是，上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在如上所述的一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (10)

  1. 一种行车盲区监测方法，其特征在于，所述行车盲区监测方法包括：
    从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
    对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
    从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的灰度变化量化值越大；
    针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
    根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
  2. 如权利要求1所述的行车盲区监测方法，其特征在于，所述从目标视频数据中确定目标图像组的步骤包括：
    从目标视频数据中获取连续N帧图像；
    根据第一公式组计算得到这N帧图像的灰度变化参考值，第一公式组为：


    其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，Tf为相邻两帧图像的 背景区域中同一像素点的灰度差异阈值，Ak(i,j)为像素点(i,j)在第k帧图像与第k-1帧图像之间的灰度差异二维值，D为背景区域，Ak为第k帧图像与第k-1帧图像的灰度变化量化值，A为灰度变化参考值；
    若这N帧图像的灰度变化参考值小于灰度变化阈值，则确定这N帧图像为目标图像组。
  3. 如权利要求1所述的行车盲区监测方法，其特征在于，所述根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果的步骤包括：
    从目标图像组后续的图像中获取连续L帧图像；
    根据第二公式组计算得到这L帧图像的第一差异参考值，第二公式组为：


    其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，Tb为障碍物监测区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Bk(i,j)为第k帧图像中的像素点(i,j)基于Tb得到的灰度差异二维值，d为障碍物监测区域，Bk为第k帧图像的障碍物监测区域中所有像素点的灰度差异二维值之和，B为第一差异参考值；
    若这L帧图像的第一差异参考值小于第一差异阈值，则判断障碍物监测区域不存在障碍物；
    若这L帧图像的第一差异参考值大于或等于第一差异阈值，则判断障碍物监测区域存在障碍物。
  4. 如权利要求3所述的行车盲区监测方法，其特征在于，在所述从目标视频数据中确定目标图像组的步骤之后还包括：
    针对背景区域中除障碍物监测区域以外的其他每个像素点，将同一像素点在这N帧图像中的灰度平均值作为该像素点的灰度标准值；
    在所述判断障碍物监测区域不存在障碍物的步骤之后还包括：
    根据第三公式组计算得到这L帧图像的第二差异参考值，第三公式组为：


    其中，fk(i,j)为第k帧图像中像素点(i,j)的灰度值，R(i,j)为像素点(i,j)的灰度标准值，TL为背景区域中同一像素点的灰度值与灰度标准值的灰度差异阈值，Ck(i,j)为第k帧图像中的像素点(i,j)基于TL得到的灰度差异二维值，D为背景区域，Ck为第k帧图像的背景区域中所有像素点的灰度差异二维值之和，C为第二差异参考值；
    若这L帧图像的第二差异参考值大于或等于第二差异阈值，则以这L帧图像的第一帧为起始帧，从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
    若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
  5. 如权利要求1所述的行车盲区监测方法，其特征在于，在所述针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值的步骤之后还包括：
    根据更新周期定期从目标视频数据中获取连续N帧图像，计算得到这N帧图像的灰度变化参考值，N等于目标图像组中的图像帧数；
    若这N帧图像的灰度变化参考值小于灰度变化阈值，则将这N帧图像作为新的目标图像组，基于新的目标图像组更新灰度标准值。
  6. 如权利要求1至5中任一项所述的行车盲区监测方法，其特征在于，在所述从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域的步骤之前还包括：
    根据车辆定位系统和高精度地图确定车辆所在车道；
    若车辆所在车道的左侧和/或右侧存在相邻车道，则将车辆对应侧的摄像头作为目标摄像头。
  7. 如权利要求6所述的行车盲区监测方法，其特征在于，背景区域包含目标摄像头拍摄的区域中的所有可行驶区域，障碍物监测区域位于目标摄像头对应的相邻车道区域中。
  8. 一种行车盲区监测装置，其特征在于，所述行车盲区监测装置包括：
    区域确定模块，用于从目标摄像头拍摄的区域中确定背景区域和障碍物监测区域，其中，背景区域包含障碍物监测区域；
    预处理模块，用于对目标摄像头采集的视频数据进行预处理，得到目标视频数据；
    图像确定模块，用于从目标视频数据中确定目标图像组，其中，目标图像组的灰度变化参考值小于灰度变化阈值，目标图像组的灰度变化参考值根据组内包含的所有相邻两帧图像的灰度变化量化值计算得到，相邻两帧图像的灰度变化量化值根据背景区域的帧间灰度变化情况确定，灰度变化越大，对应的灰度变化量化值越大；
    标准值计算模块，用于针对障碍物监测区域的每个像素点，将同一像素点在目标图像组中的灰度平均值作为该像素点的灰度标准值；
    障碍物识别模块，用于根据目标图像组后续的图像中障碍物监测区域的所有像素点的灰度值与对应的灰度标准值的差异，输出障碍物监测区域内的障碍物识别结果。
  9. 一种行车盲区监测设备，其特征在于，所述行车盲区监测设备包括处理器、存储器、以及存储在所述存储器上并可被所述处理器执行的行车盲区监测程序，其中所述行车盲区监测程序被所述处理器执行时，实现如权利要求1至7中任一项所述的行车盲区监测方法的步骤。
  10. 一种可读存储介质，其特征在于，所述可读存储介质上存储有行车盲区监测程序，其中所述行车盲区监测程序被处理器执行时，实现如权利要求1至7中任一项所述的行车盲区监测方法的步骤。
PCT/CN2024/088761 2023-09-22 2024-04-19 行车盲区监测方法、装置、设备及可读存储介质 Pending WO2025060413A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP24785958.0A EP4553794A4 (en) 2023-09-22 2024-04-19 METHOD, APPARATUS AND DEVICE FOR BLIND SPOT MONITORING DURING DRIVING, AND READABLE STORAGE HANDLE

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202311233136.2A CN117315619B (zh) 2023-09-22 2023-09-22 行车盲区监测方法、装置、设备及可读存储介质
CN202311233136.2 2023-09-22

Publications (1)

Publication Number Publication Date
WO2025060413A1 true WO2025060413A1 (zh) 2025-03-27

Family

ID=89249215

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2024/088761 Pending WO2025060413A1 (zh) 2023-09-22 2024-04-19 行车盲区监测方法、装置、设备及可读存储介质

Country Status (3)

Country Link
EP (1) EP4553794A4 (zh)
CN (1) CN117315619B (zh)
WO (1) WO2025060413A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN120431141A (zh) * 2025-07-08 2025-08-05 龙岩学院 一种基于图像融合的半挂车危险源监控方法
CN120740863A (zh) * 2025-09-01 2025-10-03 延安大学 一种用于天然气泄漏的智能检测装置及方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117315619B (zh) * 2023-09-22 2024-12-31 东风商用车有限公司 行车盲区监测方法、装置、设备及可读存储介质
CN119445296B (zh) * 2024-11-07 2025-08-29 深圳市小飞达电子有限公司 一种基于ai摄像技术的盲区精度评估方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20110075927A (ko) * 2009-12-29 2011-07-06 주식회사바텍 파노라마 영상 획득 방법 및 장치
CN111186432A (zh) * 2018-11-13 2020-05-22 杭州海康威视数字技术股份有限公司 一种车辆盲区预警的方法及装置
CN113642453A (zh) * 2021-08-11 2021-11-12 北京京东乾石科技有限公司 障碍物的检测方法、装置和系统
CN116092050A (zh) * 2022-12-29 2023-05-09 深圳市豪恩汽车电子装备股份有限公司 机动车障碍物检测方法、装置及计算机可读存储介质
CN117315619A (zh) * 2023-09-22 2023-12-29 东风商用车有限公司 行车盲区监测方法、装置、设备及可读存储介质

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7104916B2 (ja) * 2018-08-24 2022-07-22 国立大学法人岩手大学 移動物体検出装置および移動物体検出方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20110075927A (ko) * 2009-12-29 2011-07-06 주식회사바텍 파노라마 영상 획득 방법 및 장치
CN111186432A (zh) * 2018-11-13 2020-05-22 杭州海康威视数字技术股份有限公司 一种车辆盲区预警的方法及装置
CN113642453A (zh) * 2021-08-11 2021-11-12 北京京东乾石科技有限公司 障碍物的检测方法、装置和系统
CN116092050A (zh) * 2022-12-29 2023-05-09 深圳市豪恩汽车电子装备股份有限公司 机动车障碍物检测方法、装置及计算机可读存储介质
CN117315619A (zh) * 2023-09-22 2023-12-29 东风商用车有限公司 行车盲区监测方法、装置、设备及可读存储介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4553794A4 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN120431141A (zh) * 2025-07-08 2025-08-05 龙岩学院 一种基于图像融合的半挂车危险源监控方法
CN120740863A (zh) * 2025-09-01 2025-10-03 延安大学 一种用于天然气泄漏的智能检测装置及方法

Also Published As

Publication number Publication date
EP4553794A1 (en) 2025-05-14
EP4553794A4 (en) 2026-01-14
CN117315619B (zh) 2024-12-31
CN117315619A (zh) 2023-12-29

Similar Documents

Publication Publication Date Title
WO2025060413A1 (zh) 行车盲区监测方法、装置、设备及可读存储介质
CN110889351B (zh) 视频检测方法、装置、终端设备及可读存储介质
EP3798975B1 (en) Method and apparatus for detecting subject, electronic device, and computer readable storage medium
US9111169B2 (en) Method and system of identifying one or more features represented in a plurality of sensor acquired data sets
US10212397B2 (en) Abandoned object detection apparatus and method and system
US20100141806A1 (en) Moving Object Noise Elimination Processing Device and Moving Object Noise Elimination Processing Program
Atkočiūnas et al. Image processing in road traffic analysis
JP2019139729A (ja) 車線を検出するためのシステムおよび方法
CN113076851B (zh) 一种车辆违章数据的采集方法、装置以及计算机设备
CN110033425B (zh) 干扰区域检测装置及方法、电子设备
CN105046719B (zh) 一种视频监控方法及系统
WO2020140611A1 (zh) 一种基于图像处理的vin码识别方法、设备及介质
CN110647818A (zh) 一种遮挡目标物体的识别方法及装置
CN111428644A (zh) 基于深度神经网络的斑马线区域监测方法、系统及介质
US20160210759A1 (en) System and method of detecting moving objects
CN114898325A (zh) 一种车辆危险变道检测方法、装置及电子设备
WO2018068312A1 (zh) 交通异常事件检测装置及方法
CN110458029B (zh) 一种雾天环境中的车辆检测方法及装置
CN115880634A (zh) 车辆监控方法、设备及计算机可读存储介质
CN115861624B (zh) 摄像头被遮挡的检测方法、装置、设备及存储介质
CN114550060B (zh) 周界入侵识别方法、系统及电子设备
JP2002342762A (ja) 物体追跡方法
CN112927258A (zh) 一种目标跟踪方法及装置
CN115205536A (zh) 图像处理方法、装置、设备、路侧单元和自动驾驶车辆
CN119785162A (zh) 一种基于多模型融合提升事件检测正确率的方法

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2024785958

Country of ref document: EP

Effective date: 20241014

WWE Wipo information: entry into national phase

Ref document number: 2024133568

Country of ref document: RU

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

Ref document number: 24785958

Country of ref document: EP

Kind code of ref document: A1

WWP Wipo information: published in national office

Ref document number: 2024785958

Country of ref document: EP

WWG Wipo information: grant in national office

Ref document number: 2024133568

Country of ref document: RU