WO2024041464A1 - 回环路径的预测方法及装置、非易失性存储介质、处理器 - Google Patents

回环路径的预测方法及装置、非易失性存储介质、处理器 Download PDF

Info

Publication number
WO2024041464A1
WO2024041464A1 PCT/CN2023/113875 CN2023113875W WO2024041464A1 WO 2024041464 A1 WO2024041464 A1 WO 2024041464A1 CN 2023113875 W CN2023113875 W CN 2023113875W WO 2024041464 A1 WO2024041464 A1 WO 2024041464A1
Authority
WO
WIPO (PCT)
Prior art keywords
path
loopback
loop
scanning
starting position
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/CN2023/113875
Other languages
English (en)
French (fr)
Inventor
华昀峰
江腾飞
许威威
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.)
Shining 3D Technology Co Ltd
Original Assignee
Shining 3D Technology 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 Shining 3D Technology Co Ltd filed Critical Shining 3D Technology Co Ltd
Priority to EP23856554.3A priority Critical patent/EP4579189A4/en
Publication of WO2024041464A1 publication Critical patent/WO2024041464A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/38—Electronic maps specially adapted for navigation; Updating thereof
    • G01C21/3804—Creation or updating of map data
    • G01C21/3833—Creation or updating of map data characterised by the source of data
    • G01C21/3848—Data obtained from both position sensors and additional sensors
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
    • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
    • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
    • G01C21/165—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation combined with non-inertial navigation instruments
    • G01C21/1656—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation combined with non-inertial navigation instruments with passive imaging devices, e.g. cameras
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/02—Systems using the reflection of electromagnetic waves other than radio waves
    • G01S17/06—Systems determining position data of a target
    • G01S17/42—Simultaneous measurement of distance and other co-ordinates
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
    • G01S17/88—Lidar systems specially adapted for specific applications
    • G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging

Definitions

  • the present application relates to the technical field of loopback detection. Specifically, it relates to a loopback path prediction method and device, a non-volatile storage medium, and a processor.
  • Embodiments of the present application provide a loop path prediction method and device, a non-volatile storage medium, and a processor to at least solve the problem of inaccurate results of real-time positioning and map construction due to the lack of loop path prediction and loop reminder functions. technical problem.
  • a method for predicting a loop path including: during the process of instant positioning and map construction, detecting whether the scanning movement trajectory of the scanning device has not passed through it for the second time within a preset time period. The starting position on the scanning path; when it is detected that the scanning movement trajectory of the scanning device does not pass the starting position for the second time within the preset time period, the loop path is predicted.
  • prompt information is generated.
  • predicting a loopback path includes: predicting a first loopback path and/or a second loopback path, where the first loopback path is a global loopback path and the second loopback path is a local loopback path.
  • predicting the first loopback path includes: obtaining the first position of the scanning device corresponding to the current frame, and using the first position as the first endpoint of the first loopback path, where the current frame is the scanning device during the scanning process.
  • the corresponding frame when generating prompt information; obtain the starting position, and use the starting position as the second endpoint of the first loopback path; generate the first loopback path based on the first endpoint and the second endpoint of the first loopback path.
  • predicting the second loop path includes: using the first position as the first endpoint of the second loop path; obtaining the second position whose distance from the first position is the target distance, and using the second position as the second endpoint of the second loopback path; generating the second loopback path based on the first endpoint and the second endpoint of the second loopback path.
  • the second position is not on the target path, where the target path is a path in which the scanning device passes through the target position at least once, and the target position is a position other than the starting position and a distance from the starting position that is the target distance. Location.
  • all frames within the detected local loop path will be marked as loop frames; when the scanning device ends the scan, if it is detected that there is For frames other than loopback frames, prompt information is generated.
  • the above method further includes: displaying the loopback path; and/or controlling the scanning device to move along the loopback path.
  • a loop path prediction device including: a detection module configured to detect the scanning movement trajectory of the scanning device within a preset time during the process of real-time positioning and map construction. Whether the starting position on the scanning path has not been passed for the second time within a second time; the prediction module is set to predict the loop path when it is detected that the scanning movement trajectory of the scanning device has not passed the starting position for the second time within a preset time period.
  • a non-volatile storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the above loop path prediction method.
  • a processor is also provided, and the processor is configured to run a program stored in the memory, wherein the above loop path prediction method is executed when the program is running.
  • Figure 1 is a flow chart of a loop path prediction method according to an embodiment of the present application.
  • Figure 2 is a schematic diagram illustrating the difference between optimization with and without loopback according to an embodiment of the present application
  • Figure 3 is a structural diagram of a global loopback and a local loopback according to an embodiment of the present application
  • Figure 4 is a schematic diagram of a global loopback path and a local loopback path according to an embodiment of the present application
  • Figure 5 is a structural diagram of a loop path prediction device according to an embodiment of the present application.
  • an embodiment of a method for displaying a loopback path is provided. It should be noted that the steps shown in the flow chart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and ,Although a logical sequence is shown in the flowcharts, in some cases, the steps shown or described may be performed in a sequence different from that herein.
  • Figure 1 is a flow chart of a loop path prediction method according to an embodiment of the present application. The method includes the following steps:
  • Step S102 During the process of real-time positioning and map construction, it is detected whether the scanning movement trajectory of the scanning device does not pass the starting position on the scanning path for the second time within a preset time period.
  • real-time positioning and map construction is a concept: it is hoped that the robot and/or scanner will start from an unknown location in an unknown environment and repeatedly observe map features (such as, corner, pillar, etc.) to locate its own position and posture, and then incrementally construct a map based on its own position, thereby achieving the purpose of simultaneous positioning and map construction. Detects whether the scanning device passes through the start point on the scan path for the second time The starting position is the loop detection in the process of positioning and map construction. Loopback detection determines whether the robot and/or scanner has returned to a previously passed position. If a loopback is detected, it will pass the information to the backend for optimization processing.
  • map features such as, corner, pillar, etc.
  • the estimation of pose is often a recursive process, that is, the pose of the current frame is solved from the pose of the previous frame, so the error is passed on frame by frame. This is what we call cumulative error.
  • An effective way to eliminate errors is to perform loopback detection. Loop closure is a more compact and accurate constraint than the backend. This constraint can form a topologically consistent trajectory map. If closed loops can be detected and optimized, the results can be more accurate.
  • Loopback detection includes but is not limited to the following methods:
  • Loop detection through pictures The more popular loop detection method in existing simultaneous positioning and map construction systems is the method of combining feature points with bag of words (such as ORB-SLAM, VINS-Mono).
  • the method based on the bag of words is to pre-load a bag-of-words dictionary tree and instruct this preloaded dictionary tree to convert the descriptor of each local feature point in the image into a word.
  • the dictionary contains all the words. By comparing the entire image The words of the image count a word bag vector, and the distance between the word bag vectors represents the difference between the two images.
  • the inverted index method will be used to first find the key frames that have the same words as the current frame, and calculate the similarity with the current frame based on their word bag vectors, and eliminate images with insufficient similarity. frames, use the remaining key frames as candidate key frames, and sort them according to the distance of the bag of words vector from nearest to far.
  • Visual feature descriptors are highly related to the appearance of the environment. Appearance is greatly affected by lighting and changes over time, so visual feature maps tend to have a short time span.
  • Laser synchronized positioning and map construction (laser slam), loop detection for point clouds: First use Scan Context/LiDAR Iris to detect loop frames. After determining the loop frames in the historical frames, compare the loop frames with the current point cloud frames. Point cloud registration to obtain the precise pose of loop closure. The essence of loopback detection is to use the current point cloud and the historical point cloud for similarity detection. If there is a corresponding point cloud in the history with a high similarity, we will determine the historical frame as the loopback frame, and use the current point cloud and the historical frame to The precise pose is obtained through registration; due to the existence of cumulative errors, there is a certain deviation between the current moment and the historical moment calculated by the laser odometry, while the loopback detection has no cumulative error.
  • the scanner is a handheld scanner.
  • the handheld scanner In the process of using the handheld scanner for real-time positioning and map construction, it is also necessary to perform loopback detection and predict loopback when no loopback is detected. path and generate relevant prompt information.
  • the handheld scanner will A large number of frames are generated, and the three-dimensional model of the above-mentioned large scene can be reconstructed based on these large numbers of frames.
  • the handheld scanner can include cameras, inertial navigation, global positioning devices and a series of sensor elements, which can be calibrated before use.
  • Step S104 Predict the loop path when it is detected that the scanning movement trajectory of the scanning device does not pass the starting position for the second time within a preset time period.
  • loopback detection can significantly improve the quality of reconstruction. Therefore, when users scan with a lidar scanner, if the working time exceeds the preset time and it is detected that the loopback has not been completed, it is necessary to predict the loopback path. When the scanning device is not detected to pass the starting position for the second time within the preset time period, that is, when the scanning device is not detected to complete the loopback within the preset time period, the loopback path will be predicted.
  • the purpose of prompting the user to complete the loop detection is achieved, thereby achieving the technical effect of obtaining more accurate real-time positioning and map construction results, thereby solving the problem of no loop path prediction and loop closure.
  • the reminder function causes inaccurate technical problems in real-time positioning and map construction.
  • prompt information when it is detected that the scanning movement trajectory of the scanning device does not pass the starting position for the second time within a preset time period, prompt information is generated.
  • the user will be prompted: If higher reconstruction accuracy is desired, a loop needs to be formed. If a local loop is found to be formed, all frames within the loop will be marked as loop frames. When the user finally ends the scan, he still finds that there are frames that do not form loops. The user is prompted: There are some areas that do not form loops. Do you want to end the scan?
  • predicting a loopback path includes the following steps: predicting a first loopback path and/or a second loopback path, where the first loopback path is a global loopback path and the second loopback path is a local loopback path. path.
  • the global loop is a loop formed by the scanner passing through the starting position for the second time, and the local loop is formed by the scanner but the loop is not formed by passing the starting position for the second time.
  • Loopback This application can predict and guide users to complete loopback according to a global loopback path that can form a global loopback or a local loopback path that can form a local loopback.
  • predicting the first loopback path can be achieved by the following method: obtaining the first position of the scanning device corresponding to the current frame, and using the first position as the first endpoint of the first loopback path. , where the current frame is the frame corresponding to when the scanning device generates prompt information during the scanning process; obtain the starting position, and The starting position is used as the second endpoint of the first loopback path; the first loopback path is generated based on the first endpoint and the second endpoint of the first loopback path.
  • the frame generated when the user receives the reminder information is the current frame, and the current frame is used as the starting point of the global loopback path; the frame where the user's starting position is located, that is, the earliest frame (without Frames marked as loopback frames) serve as the end point of the global loopback path. Based on the starting point of the global loop path and the end point of the global loop path, a feasible path is generated for the user to choose.
  • predicting the second loop path is achieved by the following method: using the first position as the first endpoint of the second loop path, and obtaining the distance to the first position as the target distance. the second position, and use the second position as the second endpoint of the second loopback path; generate the second loopback path based on the first endpoint and the second endpoint of the second loopback path.
  • the frame generated when the user receives the reminder message is the current frame, and the current frame is used as the starting point of the local loop path; according to the current location of the user, the acquisition and current frame are set The previous frame of the distance, the time difference between the previous frame and the current frame is also greater than the set threshold, and the previous frame is not marked as a loopback frame, the previous frame is regarded as the end point of the local loopback path. Based on the starting point of the local loop path and the end point of the local loop path, a feasible path is generated for the user to choose.
  • the second position is not on the target path, where the target path is a path in which the scanning device passes through the target position at least once, and the target position is the distance between the starting position and the starting position. is a location other than the target distance.
  • the end point of the predicted local loop path should be a position that is not the starting position or a location near the starting position, and the end point of the predicted local loop path should be a frame in which no local loop has been formed, That is, the location of frames that are not marked as loopback frames. Only when the above conditions are met will a feasible local loop path be predicted for the user to choose.
  • all frames within the detected local loop path will be marked as loop frames; when the scanning device ends scanning , if a frame other than a loopback frame is detected, a prompt message is generated.
  • Figure 2 is a schematic diagram illustrating the difference between optimization with and without loopback according to an embodiment of the present application.
  • the significance of loopback detection is: related to the accuracy of the estimated trajectory and map over a long period of time; it can improve the current data Correlation with all historical data, allowing relocation using loopback detection.
  • Figure 3 is a structural diagram of a global loop and a local loop according to an embodiment of the present application.
  • the loop may be global or local.
  • the local loop pair can optimize the pose of the local frame. However, frames that do not form a loop cannot be optimized.
  • the global loop is the loop formed when the scanner passes the starting position or a position near the starting position for the second time.
  • the local loop is the loop formed when the scanner passes the same position for the second time, but the same position is not the starting position or the starting position. position near the starting position.
  • Figure 4 is a schematic diagram of a global loopback path and a local loopback path according to an embodiment of the present application.
  • the predicted loopback path is divided into a global loopback path and a local loopback path, where point b represents the point when the scanner starts to work.
  • Starting position the frame at the starting position can be called the earliest frame; point a represents the position at a set distance from the current frame.
  • the frame at point a is not marked as a loopback frame, that is, point a is not in a local loopback on the path.
  • the solid line represents the current path, and the dotted line represents the predicted loop path.
  • the loopback path may also be displayed; and/or the scanning device may be controlled to move along the loopback path.
  • the predicted feasible loop path is displayed on the display interface. Based on the predicted feasible loop path, the user can choose whether to perform loopback to obtain more accurate results for instant positioning and map construction. .
  • the scanner can be installed on an unmanned vehicle or drone to automatically scan in large scenes (such as underground parking lots, roads, stairwells, rooms, and an entire building). A community and other scenes on the order of hundreds of meters) move.
  • the previously determined loop path will be transmitted to the unmanned vehicle or drone, and the unmanned vehicle or drone will automatically plan a path based on the previously determined loop path, allowing the scanning device to move along the loop path.
  • the scanning device is controlled to move along the loopback path.
  • the scanner displays the loopback path, sends a prompt message, and after receiving a confirmation message fed back by the user, controls the scanning device to move along the loopback path.
  • Figure 5 is a structural diagram of a loop path prediction device according to an embodiment of the present application. As shown in Figure 5, the device includes:
  • the detection module 50 is configured to detect whether the scanning movement trajectory of the scanning device has not passed the starting position on the scanning path for the second time within a preset time period during the process of real-time positioning and map construction;
  • real-time positioning and map construction is a concept: Hope Robot Starting from an unknown location in an unknown environment, it locates its own position and posture through repeatedly observed map features (such as corners, pillars, etc.) during movement, and then incrementally builds a map based on its own position, thereby achieving simultaneous positioning and mapping. purpose of construction. Detecting whether the scanning device passes the starting position on the scanning path for the second time is the loopback detection in the process of positioning and map construction. Loopback detection determines whether the robot has returned to the location it previously passed. If a loopback is detected, it will pass the information to the backend for optimization.
  • the estimation of pose is often a recursive process, that is, the pose of the current frame is solved from the pose of the previous frame, so the error is passed on frame by frame. This is what we call cumulative error.
  • An effective way to eliminate errors is to perform loopback detection. Loop closure is a more compact and accurate constraint than the backend. This constraint can form a topologically consistent trajectory map. If closed loops can be detected and optimized, the results can be more accurate.
  • Loopback detection includes but is not limited to the following methods:
  • a popular loop detection method in existing simultaneous positioning and map construction systems is the method of combining feature points with bag of words.
  • the method based on the bag of words is to pre-load a bag-of-words dictionary tree and instruct this preloaded dictionary tree to convert the descriptor of each local feature point in the image into a word.
  • the dictionary contains all the words. By comparing the entire image The words of the image count a word bag vector, and the distance between the word bag vectors represents the difference between the two images.
  • the inverted index method will be used to first find the key frames that have the same words as the current frame, and calculate the similarity with the current frame based on their word bag vectors, and eliminate images with insufficient similarity.
  • Visual feature descriptors are highly related to the appearance of the environment. Appearance is greatly affected by lighting and changes over time, so visual feature maps tend to have a short time span.
  • Laser synchronized positioning and map construction, and loopback detection for point clouds First, use Scan Context/LiDAR Iris to detect loopback frames. After determining the loopback frames in the historical frames, perform point cloud registration between the loopback frames and the current point cloud frame. , to obtain the precise pose of the loop. The essence of loopback detection is to use the current point cloud and the historical point cloud for similarity detection.
  • the prediction module 52 is configured to predict the loop path when it is detected that the scanning movement trajectory of the scanning device does not pass the starting position for the second time within the preset time period;
  • real-time positioning and map construction is a concept: Hope Robot Starting from an unknown location in an unknown environment, it locates its own position and posture through repeatedly observed map features (such as corners, pillars, etc.) during movement, and then incrementally builds a map based on its own position, thereby achieving simultaneous positioning and mapping. purpose of construction. Detecting whether the scanning device passes the starting position on the scanning path for the second time is the loopback detection in the process of positioning and map construction. Loopback detection determines whether the robot has returned to the location it previously passed. If a loopback is detected, it will pass the information to the backend for optimization.
  • the estimation of pose is often a recursive process, that is, the pose of the current frame is solved from the pose of the previous frame, so the error is passed on frame by frame. This is what we call cumulative error.
  • An effective way to eliminate errors is to perform loopback detection. Loop closure is a more compact and accurate constraint than the backend. This constraint can form a topologically consistent trajectory map. If closed loops can be detected and optimized, the results can be more accurate.
  • Loopback detection includes but is not limited to the following methods:
  • a popular loop detection method in existing simultaneous positioning and map construction systems is the method of combining feature points with bag of words.
  • the method based on the bag of words is to pre-load a bag-of-words dictionary tree and instruct this preloaded dictionary tree to convert the descriptor of each local feature point in the image into a word.
  • the dictionary contains all the words. By comparing the entire image The words of the image count a word bag vector, and the distance between the word bag vectors represents the difference between the two images.
  • the inverted index method will be used to first find the key frames that have the same words as the current frame, and calculate the similarity with the current frame based on their word bag vectors, and eliminate images with insufficient similarity.
  • Visual feature descriptors are highly related to the appearance of the environment. Appearance is greatly affected by lighting and changes over time, so visual feature maps tend to have a short time span.
  • Laser synchronized positioning and map construction, and loopback detection for point clouds First, use Scan Context/LiDAR Iris to detect loopback frames. After determining the loopback frames in the historical frames, perform point cloud registration between the loopback frames and the current point cloud frame. , to obtain the precise pose of the loop. The essence of loopback detection is to use the current point cloud and the historical point cloud for similarity detection.
  • the predicted feasible loop path is displayed on the display interface. Based on the predicted feasible loop path, the user can choose whether to perform loopback to obtain more accurate results for instant positioning and map construction. .
  • Embodiments of the present application also provide a non-volatile storage medium.
  • the non-volatile storage medium includes a stored program. When the program is running, the device where the storage medium is located is controlled to execute the above loop path prediction method.
  • a program for a non-volatile storage medium to perform the following functions: during the process of instant positioning and map construction, detect whether the scanning movement trajectory of the scanning device has not passed the starting position on the scanning path for the second time within a preset time period; When it is detected that the scanning movement trajectory of the scanning device does not pass the starting position for the second time within the preset time period, the loop path is predicted.
  • An embodiment of the present application also provides a processor, which is configured to run a program stored in the memory, wherein the above loop path prediction method is executed when the program is running.
  • the processor is used to run programs that perform the following functions: during the process of real-time positioning and map construction, detect whether the scanning movement trajectory of the scanning device does not pass the starting position on the scanning path for the second time within a preset time; when detecting If the scanning movement trajectory of the scanning device does not pass the starting position for the second time within the preset time period, the loop path is predicted.
  • the disclosed technical content can be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units may be a logical functional division. In actual implementation, there may be other division methods.
  • multiple units or components may be combined or may be Integrated into another system, or some features can be ignored, or not implemented.
  • the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the units or modules may be in electrical or other forms.
  • the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present application can be integrated into one processing unit, each unit can exist physically alone, or two or more units can be integrated into one unit.
  • the above integrated units can be implemented in the form of hardware or software functional units.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, Can be stored in a computer-readable storage medium.
  • the technical solution of the present application is essentially or contributes to the relevant technology, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, It includes several instructions to cause a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application.
  • the aforementioned storage media include: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program code. .
  • the solutions provided by the embodiments of the present application can be applied to the technical field of loopback detection.
  • it is used to detect whether the scanning movement trajectory of the scanning device has not been detected for the second time within a preset time period.
  • the starting position on the scanning path has been passed twice; when it is detected that the scanning movement trajectory of the scanning device has not passed the starting position for the second time within the preset time period, the method of predicting the loop path is to predict the loop path and generate prompt information. , achieves the purpose of prompting the user to complete loopback detection, thereby achieving more accurate real-time positioning and map construction results.

Landscapes

  • Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Automation & Control Theory (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Navigation (AREA)
  • Image Analysis (AREA)
  • Traffic Control Systems (AREA)

Abstract

本申请公开了一种回环路径的预测方法及装置、非易失性存储介质、处理器。其中，该方法包括：在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。

Description

回环路径的预测方法及装置、非易失性存储介质、处理器 技术领域
本申请涉及回环检测技术领域，具体而言，涉及一种回环路径的预测方法及装置、非易失性存储介质、处理器。
背景技术
在进行即时定位与地图构建的过程中，当前帧位姿的约束是根据上一帧计算出来的，由于计算的位姿存在误差，在建图的过程中随着误差不断累积导致构建的地图容易出现漂移，因此需要通过回环检测判断扫描仪在建图过程中是否再次经过同一位置，若经过同一位置则进行回环矫正，以减小建图的偏差。但是，用户在实际应用中不会有意的去回环扫描，如果无法完成回环扫描，则会导致即时定位与地图构建的结果不准确。
针对上述的问题，目前尚未提出有效的解决方案。
发明内容
本申请实施例提供了一种回环路径的预测方法及装置、非易失性存储介质、处理器，以至少解决由于没有回环路径预测以及回环提醒功能造成的即时定位与地图构建的结果不准确的技术问题。
根据本申请实施例的一个方面，提供了一种回环路径的预测方法，包括：在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。
可选地，在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，生成提示信息。
可选地，预测回环路径，包括：预测第一回环路径和/或第二回环路径，其中，第一回环路径为全局回环路径，第二回环路径为局部回环路径。
可选地，预测第一回环路径，包括：获取当前帧对应的扫描设备的第一位置，并将第一位置作为第一回环路径的第一端点，其中，当前帧为扫描设备在扫描过程中生成提示信息时对应的帧；获取起始位置，并将起始位置作为第一回环路径的第二端点；根据第一回环路径的第一端点和第二端点，生成第一回环路径。
可选地，预测第二回环路径，包括：将第一位置作为第二回环路径的第一端点；获取与第一位置之间的距离为目标距离的第二位置，并将第二位置作为第二回环路径的第二端点；根据第二回环路径的第一端点和第二端点，生成第二回环路径。
可选地，第二位置不在目标路径上，其中，目标路径为扫描设备至少一次经过目标位置的路径，目标位置为除起始位置及与起始位置之间的距离为目标距离的位置以外的位置。
可选地，在进行即时定位与地图构建的过程中，如果检测到生成局部回环路径，将检测到的局部回环路径内的所有帧标记为回环帧；在扫描设备结束扫描时，如果检测到存在除回环帧以外的帧，生成提示信息。
可选地，预测回环路径之后，上述方法还包括：展示回环路径；和/或，控制扫描设备按照回环路径移动。
根据本申请实施例的另一方面，还提供了一种回环路径的预测装置，包括：检测模块，设置为在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；预测模块，设置为在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。
根据本申请实施例的再一方面，还提供了一种非易失性存储介质，存储介质包括存储的程序，其中，程序运行时控制存储介质所在的设备执行以上的回环路径的预测方法。
根据本申请实施例的再一方面，还提供了一种处理器，处理器设置为运行存储在存储器中的程序，其中，程序运行时执行以上的回环路径的预测方法。
在本申请实施例中，采用在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径的方式，通过预测回环路径与生成提示信息，达到了提示用户完成回环检测的目的，从而实现了获取更加精准的即时定位与地图构建结果的技术效果，进而解决了由于没有回环路径预测以及回环提醒功能造成的即时定位与地图构建的结果不准确技术问题。
附图说明
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1是根据本申请实施例的一种回环路径的预测方法的流程图；
图2是根据本申请实施例的一种有无回环优化的区别的示意图；
图3是根据本申请实施例的一种全局回环和局部回环的结构图；
图4是根据本申请实施例的一种全局回环路径和局部回环路径的示意图；
图5是根据本申请实施例的一种回环路径的预测装置的结构图。
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的实施例，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做出创造性劳动前提下所获得的所有其他实施例，都应当属于本申请保护的范围。
需要说明的是，本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本申请的实施例能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，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、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
根据本申请实施例，提供了一种回环路径的展示方法的实施例，需要说明的是，在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行，并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
图1是根据本申请实施例的一种回环路径的预测方法的流程图，该方法包括如下步骤：
步骤S102，在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置。
根据本申请的一个可选的实施例，即时定位与地图构建是一种概念：希望机器人和/或扫描仪从未知环境的未知地点出发，在运动过程中通过重复观测到的地图特征(比如，墙角，柱子等)定位自身位置和姿态，再根据自身位置增量式的构建地图，从而达到同时定位和地图构建的目的。检测扫描设备是否第二次经过扫描路径上的起 始位置即为定位与地图构建的过程中的回环检测。回环检测判断机器人和/或扫描仪是否回到了先前经过的位置，如果检测到回环，它会把信息传递给后端进行优化处理。在视觉同步定位与地图构建中，位姿的估计往往是一个递推的过程，即由上一帧位姿解算当前帧位姿，因此其中的误差便这样一帧一帧的传递下去，也就是我们所说的累积误差。一个消除误差有效的办法是进行回环检测。回环是一个比后端更加紧凑、准确的约束，这一约束条件可以形成一个拓扑一致的轨迹地图。如果能够检测到闭环，并对其优化，就可以让结果更加准确。
回环检测包括但不限于以下方法：
1.通过图片检测回环：现有的同步定位与地图构建系统中比较流行的回环检测方法是特征点结合词袋的方法(如ORB-SLAM，VINS-Mono)。基于词袋的方法是预先加载一个词袋字典树，通知这个预加载的字典树将图像中的每一局部特征点的描述子转换为一个单词，字典里包含着所有的单词，通过对整张图像的单词统计一个词袋向量，词袋向量间的距离即代表了两张图像之间的差异性。在图像检索的过程中，会利用倒排索引的方法，先找出与当前帧拥有相同单词的关键帧，并根据它们的词袋向量计算与当前帧的相似度，剔除相似度不够高的图像帧，将剩下的关键帧作为候选关键帧，按照词袋向量距离由近到远排序。视觉特征描述子与环境外观高度相关。外观受光照影响很大，且随时间变化，因此视觉特征地图往往具有较短的时效。
2.激光同步定位与地图构建(激光slam)，针对点云做回环检测：首先使用Scan Context/LiDAR Iris进行回环帧检测，确定历史帧中的回环帧后，将回环帧与当前点云帧进行点云配准，获取回环精确位姿。回环检测的本质是利用当前点云和历史点云做相似度检测，如果历史中有对应的点云相似度较高，我们就把这个历史帧确定为回环帧，用当前点云和历史帧去做配准得到精确位姿；由于累计误差的存在，激光里程计连续下来求得本时刻与那个历史时刻的位姿存在一定的偏差，而回环检测没有累计误差。
3.用全球定位系统辅助检测回环：根据全球定位系统提供的信息，判断当前位置与之前帧i的位置之间的距离，如果距离小于设置的阈值，就将当前帧与第i帧做相似度检测，如果相似度大于一定阈值，就把第i帧确定为回环帧。
作为本申请的另一个可选的实施例，扫描仪为手持扫描仪，在利用手持扫描仪进行即时定位与地图构建的过程中，同样有必要进行回环检测，并在未检测到回环时预测回环路径并生成相关提示信息。用户在利用手持扫描仪在大场景中(如地下停车场、马路、楼梯间、房间、一整栋建筑物，一个小区等百米数量级的场景)进行扫描时，扫描的过程中手持扫描仪会产生很多数量的帧，依据这些很多数量的帧可以重建上述大场景的三维模型。在利用手持扫描仪进行即时定位与地图构建的过程中，用户可以 在手持扫描仪中添加全球定位装置，也可以利用用户手机或者车辆中安装的全球定位系统，与手持扫描仪进行信号传输。用户可以通过走路或者车辆在大场景中移动。
其中，手持扫描仪内可以包括相机、惯导、全球定位装置以及一系列传感器元件，在手持扫描仪使用之前可以进行标定。
步骤S104，在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。
根据本申请的另一个可选的实施例，进行即时定位与地图构建的过程中，如果场所较大，回环检测可以显著提高重建的质量。所以用户在使用激光雷达扫描仪扫描时，如果工作时长超过预设时长，检测到还没有完成回环，就有必要预测回环的路径。在预设时长内没有检测到扫描设备第二次经过起始位置时，即预设时长内没有检测到扫描设备完成回环时，会对回环路径进行预测。
根据上述步骤，通过预测回环路径与生成提示信息，达到了提示用户完成回环检测的目的，从而实现了获取更加精准的即时定位与地图构建结果的技术效果，进而解决了由于没有回环路径预测以及回环提醒功能造成的即时定位与地图构建的结果不准确技术问题。
在本申请的一些可选的实施例，在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，生成提示信息。
作为本申请的另一个可选的实施例，如果发现扫描时间超过预设时长后，没有检测到任何回环，会提示用户：如果想要较高的重建精度，需要形成回环。如果发现形成局部回环，将回环内的所有帧标记为回环帧。用户在最后结束扫描时还发现存在没有形成回环的帧，就提示用户：有部分区域没有形成回环，是否结束扫描？
根据本申请的一个可选的实施例，预测回环路径，包括以下步骤：预测第一回环路径和/或第二回环路径，其中，第一回环路径为全局回环路径，第二回环路径为局部回环路径。
根据本申请的另一个可选的实施例，全局回环为扫描仪第二次经过起始位置所形成的回环，局部回环为扫描仪形成回环但是该回环不是第二次经过起始位置所形成的回环。本申请可以预测并指导用户按照可以形成全局回环的全局回环路径或者可以形成局部回环的局部回环路径完成回环。
在本申请的一些可选的实施例，预测第一回环路径，可以通过以下方法实现：获取当前帧对应的扫描设备的第一位置，并将第一位置作为第一回环路径的第一端点，其中，当前帧为扫描设备在扫描过程中生成提示信息时对应的帧；获取起始位置，并 将起始位置作为第一回环路径的第二端点；根据第一回环路径的第一端点和第二端点，生成第一回环路径。
作为本申请的一个可选的实施例，用户收到提醒信息时所生成的帧为当前帧，将当前帧作为全局回环路径的起点；用户起始位置所在的帧，也即最早的帧(没有被标记为回环帧的帧)作为全局回环路径的终点。根据全局回环路径的起点和全局回环路径的终点，生成一条可行的路径供用户选择。
在本申请的一些可选的实施例中，预测第二回环路径，通过以下方法实现：将第一位置作为第二回环路径的第一端点，获取与第一位置之间的距离为目标距离的第二位置，并将第二位置作为第二回环路径的第二端点；根据第二回环路径的第一端点和第二端点，生成第二回环路径。
作为本申请的另一个可选的实施例，用户收到提醒信息时所生成的帧为当前帧，将当前帧作为局部回环路径的起点；根据用户所在的当前位置，获取与当前帧为设定距离的先前帧，该先前帧与当前帧的时间差同样也大于设定阈值，且该先前帧没有被标记为回环帧，将该先前帧作为局部回环路径的终点。根据局部回环路径的起点和局部回环路径的终点，生成一条可行的路径供用户选择。
根据本申请的一个可选的实施例，第二位置不在目标路径上，其中，目标路径为扫描设备至少一次经过目标位置的路径，目标位置为除起始位置及与起始位置之间的距离为目标距离的位置以外的位置。
作为本申请的一个可选的实施例，预测的局部回环路径的终点应为不是起始位置或起始位置的附近位置，而且预测的局部回环路径的终点应为没有形成过局部回环的帧，也即没有被标记为回环帧的帧所在的位置。满足以上条件才会预测出一条可行的局部回环路径供用户选择。
在本申请的一些可选的实施例中，定位与地图构建的过程中，如果检测到生成局部回环路径，将检测到的局部回环路径内的所有帧标记为回环帧；在扫描设备结束扫描时，如果检测到存在除回环帧以外的帧，生成提示信息。
作为本申请的另一个可选的实施例，在进行即时定位与地图构建的过程中，如果生成了局部回环，则会将局部回环内的所有帧标记为回环帧，当用户在最后结束扫描时还发现存在有没有形成回环的帧，也即形成了多个局部回环但是还没有生成全局回环，就会提示用户：有部分区域没有形成回环，是否结束扫描？当用户想要更加精准的即时定位与地图构建结果时，用户可以按照生成的回环预测路径完成回环。
图2是根据本申请实施例的一种有无回环优化的区别的示意图，如图2所示，回环检测的意义为：关系到估计的轨迹和地图在长时间下的正确性；能够提高当前数据 与所有历史数据的关联，从而可以利用回环检测进行重定位。
图3是根据本申请实施例的一种全局回环和局部回环的结构图，如图3所示，回环可能是全局的，也可能是局部的，局部的回环对可以优化局部帧的位姿，但是不能优化没有形成回环的帧。全局回环为扫描仪第二次经过起始位置或起始位置附近的位置所形成的回环，局部回环为扫描仪第二次经过相同位置所形成的回环，但该相同位置不是起始位置或者起始位置附近的位置。
图4是根据本申请实施例的一种全局回环路径和局部回环路径的示意图，如图4所示，预测的回环路径分为全局回环路径和局部回环路径，其中b点表示扫描仪开始工作的起始位置，起始位置上的帧可以称为最早帧；a点表示与当前帧所在位置为设定距离的位置，a点上的帧没有被标记为回环帧，也即a点不在局部回环的路径上。其中，实线表示为当前的路径，虚线表示为预测的回环路径。
作为本申请的一个可选的实施例，执行步骤S104预测回环路径之后，还可以展示回环路径；和/或，控制扫描设备按照回环路径移动。
在本申请的一些可选的实施例，在显示界面展示预测出来的可行的回环路径，用户根据预测出来的可行的回环路径，可以选择是否进行回环以获得即时定位与地图构建的更准确和结果。
作为本申请的另一个可选的实施例，扫描仪可以安装于无人驾驶车辆或无人机上，自动在大场景中(如地下停车场、马路、楼梯间、房间、一整栋建筑物，一个小区等百米数量级的场景)移动。之前确定的回环路径会传输至无人驾驶车辆或无人机，无人驾驶车辆或无人机会根据之前确定的回环路径自动规划路径运行，从而使得扫描设备按照回环路径移动。
作为本申请的另一个可选的实施例，执行步骤S104预测回环路径之后，在展示回环路径的同时，控制扫描设备按照回环路径移动。
在本申请的一些可选的实施例中，执行步骤S104预测回环路径之后，扫描仪展示回环路径后，发送提示消息，当接受到用户反馈的确认消息后，控制扫描设备按照回环路径移动。
图5是根据本申请实施例的一种回环路径的预测装置的结构图，如图5所示，该装置包括：
检测模块50，设置为在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；
根据本申请的一个可选的实施例，即时定位与地图构建是一种概念：希望机器人 从未知环境的未知地点出发，在运动过程中通过重复观测到的地图特征(比如，墙角，柱子等)定位自身位置和姿态，再根据自身位置增量式的构建地图，从而达到同时定位和地图构建的目的。检测扫描设备是否第二次经过扫描路径上的起始位置即为定位与地图构建的过程中的回环检测。回环检测判断机器人是否回到了先前经过的位置，如果检测到回环，它会把信息传递给后端进行优化处理。在视觉同步定位与地图构建中，位姿的估计往往是一个递推的过程，即由上一帧位姿解算当前帧位姿，因此其中的误差便这样一帧一帧的传递下去，也就是我们所说的累积误差。一个消除误差有效的办法是进行回环检测。回环是一个比后端更加紧凑、准确的约束，这一约束条件可以形成一个拓扑一致的轨迹地图。如果能够检测到闭环，并对其优化，就可以让结果更加准确。
回环检测包括但不限于以下方法：
1.通过图片检测回环：现有的同步定位与地图构建系统中比较流行的回环检测方法是特征点结合词袋的方法。基于词袋的方法是预先加载一个词袋字典树，通知这个预加载的字典树将图像中的每一局部特征点的描述子转换为一个单词，字典里包含着所有的单词，通过对整张图像的单词统计一个词袋向量，词袋向量间的距离即代表了两张图像之间的差异性。在图像检索的过程中，会利用倒排索引的方法，先找出与当前帧拥有相同单词的关键帧，并根据它们的词袋向量计算与当前帧的相似度，剔除相似度不够高的图像帧，将剩下的关键帧作为候选关键帧，按照词袋向量距离由近到远排序。视觉特征描述子与环境外观高度相关。外观受光照影响很大，且随时间变化，因此视觉特征地图往往具有较短的时效。
2.激光同步定位与地图构建，针对点云做回环检测：首先使用Scan Context/LiDAR Iris进行回环帧检测，确定历史帧中的回环帧后，将回环帧与当前点云帧进行点云配准，获取回环精确位姿。回环检测的本质是利用当前点云和历史点云做相似度检测，如果历史中有对应的点云相似度较高，我们就把这个历史帧确定为回环帧，用当前点云和历史帧去做配准得到精确位姿；由于累计误差的存在，激光里程计连续下来求得本时刻与那个历史时刻的位姿存在一定的偏差，而回环检测没有累计误差。
3.用全球定位系统辅助检测回环：根据全球定位系统提供的信息，判断当前位置与之前帧i的位置之间的距离，如果距离小于设置的阈值，就将当前帧与第i帧做相似度检测，如果相似度大于一定阈值，就把第i帧确定为回环帧。
预测模块52，设置为在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径；
根据本申请的一个可选的实施例，即时定位与地图构建是一种概念：希望机器人 从未知环境的未知地点出发，在运动过程中通过重复观测到的地图特征(比如，墙角，柱子等)定位自身位置和姿态，再根据自身位置增量式的构建地图，从而达到同时定位和地图构建的目的。检测扫描设备是否第二次经过扫描路径上的起始位置即为定位与地图构建的过程中的回环检测。回环检测判断机器人是否回到了先前经过的位置，如果检测到回环，它会把信息传递给后端进行优化处理。在视觉同步定位与地图构建中，位姿的估计往往是一个递推的过程，即由上一帧位姿解算当前帧位姿，因此其中的误差便这样一帧一帧的传递下去，也就是我们所说的累积误差。一个消除误差有效的办法是进行回环检测。回环是一个比后端更加紧凑、准确的约束，这一约束条件可以形成一个拓扑一致的轨迹地图。如果能够检测到闭环，并对其优化，就可以让结果更加准确。
回环检测包括但不限于以下方法：
1.通过图片检测回环：现有的同步定位与地图构建系统中比较流行的回环检测方法是特征点结合词袋的方法。基于词袋的方法是预先加载一个词袋字典树，通知这个预加载的字典树将图像中的每一局部特征点的描述子转换为一个单词，字典里包含着所有的单词，通过对整张图像的单词统计一个词袋向量，词袋向量间的距离即代表了两张图像之间的差异性。在图像检索的过程中，会利用倒排索引的方法，先找出与当前帧拥有相同单词的关键帧，并根据它们的词袋向量计算与当前帧的相似度，剔除相似度不够高的图像帧，将剩下的关键帧作为候选关键帧，按照词袋向量距离由近到远排序。视觉特征描述子与环境外观高度相关。外观受光照影响很大，且随时间变化，因此视觉特征地图往往具有较短的时效。
2.激光同步定位与地图构建，针对点云做回环检测：首先使用Scan Context/LiDAR Iris进行回环帧检测，确定历史帧中的回环帧后，将回环帧与当前点云帧进行点云配准，获取回环精确位姿。回环检测的本质是利用当前点云和历史点云做相似度检测，如果历史中有对应的点云相似度较高，我们就把这个历史帧确定为回环帧，用当前点云和历史帧去做配准得到精确位姿；由于累计误差的存在，激光里程计连续下来求得本时刻与那个历史时刻的位姿存在一定的偏差，而回环检测没有累计误差。
3.用全球定位系统辅助检测回环：根据全球定位系统提供的信息，判断当前位置与之前帧i的位置之间的距离，如果距离小于设置的阈值，就将当前帧与第i帧做相似度检测，如果相似度大于一定阈值，就把第i帧确定为回环帧。
在本申请的一些可选的实施例，在显示界面展示预测出来的可行的回环路径，用户根据预测出来的可行的回环路径，可以选择是否进行回环以获得即时定位与地图构建的更准确和结果。
需要说明的是，图5所示实施例的优选实施方式可以参见图1所示实施例的相关描述，此处不再赘述。
本申请实施例还提供一种非易失性存储介质，非易失性存储介质包括存储的程序，其中，程序运行时控制存储介质所在的设备执行以上的回环路径的预测方法。
非易失性存储介质执行以下功能的程序：在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。
本申请实施例还提供了一种处理器，处理器设置为运行存储在存储器中的程序，其中，程序运行时执行以上的回环路径的预测方法。
处理器用于运行执行以下功能的程序：在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径。
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。
在本申请的上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。
在本申请所提供的几个实施例中，应该理解到，所揭露的技术内容，可通过其它的方式实现。其中，以上所描述的装置实施例仅仅是示意性的，例如所述单元的划分，可以为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，单元或模块的间接耦合或通信连接，可以是电性或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本申请各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
所述集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时， 可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对相关技术做出贡献的部分或者该技术方案的全部或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一个存储介质中，包括若干指令用以使得一台计算机设备(可为个人计算机、服务器或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、移动硬盘、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述仅是本申请的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本申请原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本申请的保护范围。
工业实用性
本申请实施例提供的方案可应用于回环检测技术领域，在本申请实施例中，采用在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；在检测到扫描设备的扫描移动轨迹在预设时长内未第二次经过起始位置的情况下，预测回环路径的方式，通过预测回环路径与生成提示信息，达到了提示用户完成回环检测的目的，从而实现了获取更加精准的即时定位与地图构建结果。

Claims (11)

  1. 一种回环路径的预测方法，包括：
    在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；
    在检测到所述扫描设备的扫描移动轨迹在预设时长内未第二次经过所述起始位置的情况下，预测回环路径。
  2. 根据权利要求1所述的方法，其中，所述方法还包括：在检测到所述扫描设备的扫描移动轨迹在所述预设时长内未第二次经过所述起始位置的情况下，生成提示信息。
  3. 根据权利要求2所述的方法，其中，预测回环路径，包括：
    预测第一回环路径和/或第二回环路径，其中，所述第一回环路径为全局回环路径，所述第二回环路径为局部回环路径。
  4. 根据权利要求3所述的方法，其中，预测第一回环路径，包括：
    获取当前帧对应的扫描设备的第一位置，并将所述第一位置作为所述第一回环路径的第一端点，其中，所述当前帧为所述扫描设备在扫描过程中生成所述提示信息时对应的帧；
    获取所述起始位置，并将所述起始位置作为所述第一回环路径的第二端点；
    根据所述第一回环路径的第一端点和第二端点，生成所述第一回环路径。
  5. 根据权利要求4所述的方法，其中，预测第二回环路径，包括：
    将所述第一位置作为所述第二回环路径的第一端点；
    获取与所述第一位置之间的距离为目标距离的第二位置，并将所述第二位置作为所述第二回环路径的第二端点；
    根据所述第二回环路径的第一端点和第二端点，生成所述第二回环路径。
  6. 根据权利要求5所述的方法，其中，
    所述第二位置不在目标路径上，其中，所述目标路径为所述扫描设备至少一次经过目标位置的路径，所述目标位置为除所述起始位置及与所述起始位置之间的距离为所述目标距离的位置以外的位置。
  7. 根据权利要求3所述的方法，其中，所述方法还包括：
    在进行所述即时定位与地图构建的过程中，如果检测到生成所述局部回环路径，将检测到的所述局部回环路径内的所有帧标记为回环帧；
    在所述扫描设备结束扫描时，如果检测到存在除所述回环帧以外的帧，生成所述提示信息。
  8. 根据权利要求1至7中任意一项所述的方法，其中，预测回环路径之后，所述方法还包括：
    展示所述回环路径；和/或
    控制所述扫描设备按照所述回环路径移动。
  9. 一种回环路径的预测装置，包括：
    检测模块，设置为在进行即时定位与地图构建的过程中，检测扫描设备的扫描移动轨迹在预设时长内是否未第二次经过扫描路径上的起始位置；
    预测模块，设置为在检测到所述扫描设备的扫描移动轨迹在预设时长内未第二次经过所述起始位置的情况下，预测回环路径。
  10. 一种非易失性存储介质，所述非易失性存储介质包括存储的程序，其中，在所述程序运行时控制所述非易失性存储介质所在设备执行权利要求1至8中任意一项所述的回环路径的预测方法。
  11. 一种处理器，所述处理器设置为运行存储在存储器中的程序，其中，所述程序运行时执行权利要求1至8中任意一项所述的回环路径的预测方法。
PCT/CN2023/113875 2022-08-22 2023-08-18 回环路径的预测方法及装置、非易失性存储介质、处理器 Ceased WO2024041464A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
EP23856554.3A EP4579189A4 (en) 2022-08-22 2023-08-18 METHOD AND APPARATUS FOR LOOP-BASED TRACK PREDICTION WITH RETURN, NON-VOLATILE STORAGE MEDIA, AND PROCESSOR

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202211007262.1 2022-08-22
CN202211007262.1A CN115420275B (zh) 2022-08-22 2022-08-22 回环路径的预测方法及装置、非易失性存储介质、处理器

Publications (1)

Publication Number Publication Date
WO2024041464A1 true WO2024041464A1 (zh) 2024-02-29

Family

ID=84197968

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/113875 Ceased WO2024041464A1 (zh) 2022-08-22 2023-08-18 回环路径的预测方法及装置、非易失性存储介质、处理器

Country Status (3)

Country Link
EP (1) EP4579189A4 (zh)
CN (1) CN115420275B (zh)
WO (1) WO2024041464A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118938166A (zh) * 2024-10-12 2024-11-12 山东大学 基于点云几何信息的回环检测方法、系统、介质及设备

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115420275B (zh) * 2022-08-22 2026-03-17 先临三维科技股份有限公司 回环路径的预测方法及装置、非易失性存储介质、处理器
CN115661255B (zh) * 2022-12-15 2023-04-28 中国科学技术大学 一种激光slam回环检测与校正方法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017146952A (ja) * 2016-02-15 2017-08-24 キヤノン株式会社 情報処理装置、情報処理装置の制御方法およびプログラム
CN111862162A (zh) * 2020-07-31 2020-10-30 湖北亿咖通科技有限公司 回环检测方法及系统、可读存储介质、电子设备
CN112955901A (zh) * 2018-12-23 2021-06-11 三星电子株式会社 用于执行回环检测的方法和装置
CN113791616A (zh) * 2021-08-25 2021-12-14 深圳优地科技有限公司 路径规划方法、装置、机器人和存储介质
CN114034299A (zh) * 2021-11-08 2022-02-11 中南大学 一种基于主动激光slam的导航系统
CN114674311A (zh) * 2022-03-22 2022-06-28 中国矿业大学 一种室内定位与建图方法及系统
CN115420275A (zh) * 2022-08-22 2022-12-02 先临三维科技股份有限公司 回环路径的预测方法及装置、非易失性存储介质、处理器

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107958482B (zh) * 2016-10-17 2021-06-18 杭州海康威视数字技术股份有限公司 一种三维场景模型构建方法及装置
CN114689062B (zh) * 2020-12-28 2025-09-19 深圳市普渡科技有限公司 地图构建方法、装置、计算机设备及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017146952A (ja) * 2016-02-15 2017-08-24 キヤノン株式会社 情報処理装置、情報処理装置の制御方法およびプログラム
CN112955901A (zh) * 2018-12-23 2021-06-11 三星电子株式会社 用于执行回环检测的方法和装置
CN111862162A (zh) * 2020-07-31 2020-10-30 湖北亿咖通科技有限公司 回环检测方法及系统、可读存储介质、电子设备
CN113791616A (zh) * 2021-08-25 2021-12-14 深圳优地科技有限公司 路径规划方法、装置、机器人和存储介质
CN114034299A (zh) * 2021-11-08 2022-02-11 中南大学 一种基于主动激光slam的导航系统
CN114674311A (zh) * 2022-03-22 2022-06-28 中国矿业大学 一种室内定位与建图方法及系统
CN115420275A (zh) * 2022-08-22 2022-12-02 先临三维科技股份有限公司 回环路径的预测方法及装置、非易失性存储介质、处理器

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4579189A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118938166A (zh) * 2024-10-12 2024-11-12 山东大学 基于点云几何信息的回环检测方法、系统、介质及设备

Also Published As

Publication number Publication date
CN115420275A (zh) 2022-12-02
CN115420275B (zh) 2026-03-17
EP4579189A4 (en) 2025-11-26
EP4579189A1 (en) 2025-07-02

Similar Documents

Publication Publication Date Title
WO2024041464A1 (zh) 回环路径的预测方法及装置、非易失性存储介质、处理器
CN111445526B (zh) 一种图像帧之间位姿的估计方法、估计装置和存储介质
EP3777119B1 (en) Stabilizing video to reduce camera and face movement
CN107990899B (zh) 一种基于slam的定位方法和系统
CN111094895B (zh) 用于在预构建的视觉地图中进行鲁棒自重新定位的系统和方法
KR102410879B1 (ko) 포지셔닝 정보를 획득하는 방법, 장치 및 매체
CN113378605B (zh) 多源信息融合方法及装置、电子设备和存储介质
JPWO2017168899A1 (ja) 情報処理方法および情報処理装置
CN107784671A (zh) 一种用于视觉即时定位与建图的方法与系统
CN111105695B (zh) 地图制作方法、装置、电子设备及计算机可读存储介质
US12254041B2 (en) Position recognition method and system based on visual information processing
WO2022193456A1 (zh) 目标跟踪方法及装置、电子设备和存储介质
CN110349212A (zh) 即时定位与地图构建的优化方法及装置、介质和电子设备
CN113194253A (zh) 去除图像反光的拍摄方法、装置和电子设备
WO2020114214A1 (zh) 导盲方法和装置，存储介质和电子设备
CN113465600B (zh) 一种导航方法、装置及电子设备和存储介质
CN110728756A (zh) 一种基于增强现实的远程指导的方法与设备
CN114241146B (zh) 一种构建地图的方法、装置、电子设备及存储介质
Einizinab et al. Camera Pose Refinement for Precise BIM Alignment in Mixed Reality Visualization
US10735902B1 (en) Method and computer program for taking action based on determined movement path of mobile devices
Shuai et al. Multi-sensor fusion for autonomous positioning of indoor robots
TWI768724B (zh) 立體空間中定位方法與定位系統
CN114730457B (zh) 广角白平衡
TWI788217B (zh) 立體空間中定位方法與定位系統
CN117974770A (zh) 一种数据处理的方法、装置、电子设备及介质

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: 23856554

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2023856554

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2023856554

Country of ref document: EP

Effective date: 20250324

WWP Wipo information: published in national office

Ref document number: 2023856554

Country of ref document: EP