Obstacle detection method and device and storage medium

Abstract

The invention discloses an obstacle detection method and device and a storage medium, and belongs to the technical field of vehicle control. The method comprises the following steps: collecting information of a reflection signal corresponding to a laser pulse emitted by a radar point cloud sensor; generating point cloud data of the obstacle based on the information of the reflection signal, wherein the point cloud data comprises space coordinates of the reflection point; calculating the space coordinate of the center point of the obstacle based on the space coordinate of the reflection point; generating an initial obstacle prediction result based on the space coordinate of the center point of the obstacle and the space coordinate of the reflection point; obtaining a visibility detection result of the side surface based on the overlook length of the marking box of the obstacle and the point cloud data of the side surface of the obstacle; in response to the side being visible relative to the radar point cloud sensor, calculating spatial coordinates of a center point of the side based on the point cloud data of the side of the obstacle; and correcting the initial obstacle prediction result based on the space coordinates of the central points of the side surfaces to obtain a final obstacle prediction result. And the accuracy of obstacle position detection is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of vehicle control, and particularly to a method, device and storage medium for detecting obstacles. Background Art

[0002] In the technical field of autonomous driving, it is necessary to detect the type and position of obstacles on the road where the vehicle is located, so as to conduct a risk assessment on the obstacles and prevent the vehicle from colliding with the obstacles. Among them, the obstacles may be objects, pedestrians or other vehicles on the road. In the related art, on-vehicle lidar sensors are used to detect obstacles. In the related art, the detection of obstacles lacks accuracy, and it is difficult to detect when the center of the obstacle is blocked. Therefore, how to improve the accuracy of obstacle detection and achieve obstacle detection in various situations is very important for preventing the vehicle from colliding with obstacles and improving the safety of vehicle driving. Summary of the Invention

[0003] The embodiments of the present application provide a method, device and storage medium for detecting obstacles, which can be used to improve the safety of vehicle driving to a certain extent. The technical solutions are as follows:

[0004] On the one hand, the embodiments of the present application provide a method for detecting obstacles, and the method includes:

[0005] Collect information on the reflection signals corresponding to the laser pulses emitted by the radar point cloud sensor;

[0006] Generate point cloud data of the obstacle based on the information of the reflection signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflection signal;

[0007] Calculate the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle;

[0008] Generate an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points, where the initial obstacle prediction result indicates the initial position and shape of the obstacle;

[0009] Obtain the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, where the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located;

[0010] In response to obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, calculate the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle;

[0011] Based on the spatial coordinates of the center point of the side surface, correct the initial obstacle prediction result to obtain a final obstacle prediction result, where the final obstacle prediction result indicates the final position of the obstacle.

[0012] On the other hand, a device for detecting an obstacle is provided, and the device includes:

[0013] An acquisition module, configured to acquire information on a reflection signal corresponding to a laser pulse emitted by a radar point cloud sensor;

[0014] A first generation module, configured to generate point cloud data of an obstacle based on the information on the reflection signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflection signal;

[0015] A first calculation module, configured to calculate the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle;

[0016] A second generation module, configured to generate an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points, where the initial obstacle prediction result indicates the initial position and shape of the obstacle;

[0017] An acquisition module, configured to obtain a visibility detection result of the side surface based on the top-down length of the annotation box of the obstacle and the point cloud data of the side surface of the obstacle, where the visibility detection result indicates whether the side surface is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located;

[0018] A second calculation module, configured to, in response to obtaining the visibility detection result indicating that the side surface is visible relative to the radar point cloud sensor, calculate the spatial coordinates of the center point of the side surface based on the point cloud data of the side surface of the obstacle;

[0019] A correction module, configured to correct the initial obstacle prediction result based on the spatial coordinates of the center point of the side surface to obtain a final obstacle prediction result, where the final obstacle prediction result indicates the final position of the obstacle.

[0020] On the other hand, a non-transitory computer-readable storage medium is further provided, characterized in that a computer program is stored in the computer-readable storage medium, and the computer program is loaded and executed by a processor to implement the obstacle detection method described in any one of the above.

[0021] On the other hand, a computer program product is also provided. The computer program product includes computer instructions that, when executed by a processor, implement the steps of any one of the above-described obstacle detection methods.

[0022] The technical solutions provided in this application at least bring the following beneficial effects:

[0023] In this application, by collecting the information of the reflection signals corresponding to the radar point cloud sensor, the point cloud data of the obstacle is obtained, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle; then, based on the spatial coordinates of the reflection points of the obstacle, the spatial coordinates of the center point of the obstacle are calculated, which are used to generate an initial obstacle prediction result in combination with the spatial coordinates of the reflection points; the visibility detection result of the side is judged based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle; after determining that any side is visible relative to the radar point cloud sensor, based on the point cloud data of this side, the spatial coordinates of the center point of this side are calculated, which are used to correct the initial obstacle prediction result. By calculating the spatial coordinates of the center point of the visible surface, the position of the obstacle is corrected, improving the accuracy of obstacle position detection and making the process of predicting the position of the obstacle more convenient to implement. Description of the Drawings

[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following-described drawings are only some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0025] Figure 1 It is a schematic diagram of an implementation environment provided by an embodiment of this application;

[0026] Figure 2 It is a flowchart of an obstacle detection method provided by an embodiment of this application;

[0027] Figure 3 It is a schematic structural diagram of an obstacle detection device provided by an embodiment of this application. Detailed Embodiments

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the following will further describe the embodiments of this application in detail with reference to the drawings.

[0029] An embodiment of this application provides an obstacle detection method. Please refer to Figure 1 , which shows a schematic diagram of the method implementation environment provided by an embodiment of this application. The implementation environment may include: an obstacle detection system 11, a radar point cloud sensor 12, a camera 13, and a display screen 14 of the center console.

[0030] Optionally, the radar point cloud sensor 12 is located in any direction around the vehicle and can be used to emit laser pulses around the vehicle, collect the reflection signals corresponding to the laser pulses, and analyze the information of the reflection signals and send it to the obstacle detection system 11. Among them, the information of the reflection signals includes the propagation direction of the reflection signals and the duration from the emission of the laser pulses to the reception of the reflection signals.

[0031] Exemplarily, the camera 13 is installed in any direction around the vehicle and can be used to capture the environment around the vehicle and transmit the captured environment to the visual recognition device and the display screen 14 on the center console. The display screen 14 on the center console is located on the center console of the vehicle and can display the environment where the vehicle is located captured by the camera 13 and visualize the bounding box according to the data of the bounding box, that is, select the obstacles in the video of the environment where the vehicle is located. Among them, the obstacle detection system 11, the radar point cloud sensor 12, the camera 13, and the display screen 14 on the center console establish a communication connection through a wired or wireless network.

[0032] Based on the above Figure 1 shown implementation environment, an embodiment of the present application provides a method for detecting obstacles as Figure 2 shown. Taking the application of this method to the obstacle detection system as an example, this method includes steps 201 - step 207.

[0033] In step 201, the obstacle detection system collects the information of the reflection signals corresponding to the laser pulses emitted by the radar point cloud sensor.

[0034] In a possible implementation manner, the radar point cloud sensor can be installed in any direction around the vehicle, and the obstacle detection system controls the radar point cloud sensor to emit laser pulses around the vehicle. The laser pulses will be reflected after encountering obstacles, generating reflection signals. Among them, the obstacles can be other vehicles, pedestrians, and objects that hinder the vehicle's driving, etc.

[0035] Optionally, the obstacle detection system collects the information of the reflection signals corresponding to the laser pulses emitted by the radar point cloud sensor, including: the obstacle detection system collects the propagation direction of the reflection signals and the duration from the emission of the laser pulses to the reception of the reflection signals from the radar point cloud sensor.

[0036] In step 202, the obstacle detection system generates the point cloud data of the obstacle based on the information of the reflection signals. The point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflection signal.

[0037] Exemplarily, after collecting the information of the reflected signal, the obstacle detection system generates point cloud data of the obstacle based on the information of the reflected signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflected signal.

[0038] In a possible implementation manner, the obstacle detection system generates point cloud data of the obstacle based on the information of the reflected signal, including: calculating the distance between the reflection point and the radar point cloud sensor based on the duration from emitting the laser pulse to receiving the reflected signal; generating the spatial coordinates of the reflection point based on the propagation direction of the reflected signal and the distance between the reflection point and the radar point cloud sensor. Among them, the spatial coordinate system can be established in advance. For example, taking the position of the vehicle as the origin, the direction perpendicular to the road surface as the Z-axis, the direction perpendicular to the vehicle bumper as the Y-axis, and the direction parallel to the bumper as the X-axis.

[0039] Exemplarily, the method for calculating the distance between the reflection point and the radar point cloud sensor based on the duration from emitting the laser pulse to receiving the reflected signal includes, but is not limited to: calculating the product of the duration from emitting the laser pulse to receiving the reflected signal multiplied by the speed of light, and then dividing the product by 2 to obtain the calculation result, and taking the calculation result as the distance between the reflection point and the radar point cloud sensor. In a possible implementation manner, the obstacle detection system can also process the reflected signal through a target detection algorithm, and combine the camera and the visual recognition device to judge the type of the obstacle. Among them, the camera is used to photograph the environment around the vehicle.

[0040] In step 203, the obstacle detection system calculates the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle.

[0041] Optionally, after determining the spatial coordinates of the reflection points of the obstacle, the obstacle detection system performs denoising processing on the spatial coordinates of the reflection points. For example, the obstacle detection system filters the spatial coordinates of the reflection points through a filter to remove noise signals and outliers. After completing the denoising, the obstacle detection system calculates the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle, including: the obstacle detection system performs segmentation processing on the spatial coordinates of the reflection points through a clustering algorithm to determine the obstacle to which each reflection point belongs. Then, the spatial coordinates of the reflection points included in each obstacle are input into a preset first center point coordinate calculation model to obtain the spatial coordinates of the center point of the obstacle corresponding to the spatial coordinates of the reflection points of the obstacle.

[0042] In a possible implementation manner, the first center point coordinate calculation model is trained by using the point cloud data of the object with the spatial coordinates of the center point marked, and can be used to calculate the spatial coordinates of the center point of the object according to the spatial coordinates of the points on the surface of the object.

[0043] In step 204, the obstacle detection system generates an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection point, and the initial obstacle prediction result indicates the initial position and shape of the obstacle.

[0044] Optionally, after calculating the spatial coordinates of the center point of the obstacle, the obstacle detection system generates an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection point, where the initial obstacle prediction result indicates the initial position and shape of the obstacle.

[0045] In a possible implementation, the obstacle detection system generates an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection point, including: determining the initial position of the obstacle based on the spatial coordinates of the center point of the obstacle; calculating the shape of the obstacle based on the spatial coordinates of the reflection point of the obstacle.

[0046] Exemplarily, determining the initial position of the obstacle based on the spatial coordinates of the center point of the obstacle includes: using the spatial coordinates of the center point of the obstacle as the coordinates of the initial position of the obstacle. Calculating the shape of the obstacle based on the spatial coordinates of the reflection point of the obstacle includes: calculating the length, width, and height of the obstacle based on the spatial coordinates of the reflection point, so as to determine the shape of the obstacle.

[0047] Optionally, the obstacle detection system can also determine the data of the bounding box of the obstacle according to the spatial coordinates of the center point of the obstacle, the length, width, and height of the obstacle, then display the environment where the vehicle is located captured by the camera through the display screen of the vehicle center console, and visualize the data of the bounding box to frame the obstacle in the video.

[0048] In step 205, the obstacle detection system obtains a side visibility detection result based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, and the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located.

[0049] In a possible implementation, after obtaining the initial obstacle prediction result and determining the data of the bounding box of the obstacle, the obstacle detection system obtains a side visibility detection result based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, where the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located.

[0050] Exemplarily, the obstacle detection system obtains the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, including: determining the spatial coordinates of the boundary points of the side based on the point cloud data of the side of the obstacle; calculating the length of the point cloud region of the side based on the spatial coordinates of the boundary points; counting the number of points included in the side; and in response to the ratio of the top-down length of the annotation box to the length of the point cloud region of the side being greater than the ratio threshold and the number of points included in the side being greater than the number threshold, obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor.

[0051] Optionally, the obstacle detection system calculates the top-down length of the annotation box based on the data of the bounding box, segments the point cloud data of each side of the obstacle, then determines the spatial coordinates of the boundary points of the side based on the point cloud data of each side of the obstacle, and counts the number of points included in each side. After determining the spatial coordinates of the boundary points of the side, the length of the point cloud region of the side is calculated based on the spatial coordinates of the boundary points using the length calculation formula of the spatial coordinates.

[0052] In a possible implementation manner, the obstacle detection system compares the number of points included in each side with the number threshold, and then calculates the ratio of the top-down length of the annotation box to the length of the point cloud region of the side. If the ratio of the top-down length of the annotation box to the length of the point cloud region of the side is greater than the ratio threshold and the number of points included in the side is greater than the number threshold, the obstacle detection system obtains the visibility detection result indicating that the side is visible relative to the radar point cloud sensor; if at least one of the ratio of the top-down length of the annotation box to the length of the point cloud region of the side being less than or equal to the ratio threshold or the number of points included in the side being less than or equal to the number threshold is satisfied, the obstacle detection system obtains the visibility detection result indicating that the side is not visible relative to the radar point cloud sensor.

[0053] In step 206, in response to obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, the obstacle detection system calculates the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle.

[0054] Exemplarily, after obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, the obstacle detection system calculates the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle, including: the obstacle detection system inputs the spatial coordinates of the reflection points included in each side into a preset second center point coordinate calculation model to obtain the spatial coordinates of the center point of the side corresponding to the spatial coordinates of the reflection points of the side.

[0055] In a possible implementation, the second center point coordinate calculation model is trained with the point cloud data of an object with the spatial coordinates of the center point on the side marked, and can be used to calculate the spatial coordinates of the center point on the side of the object according to the spatial coordinates of the points on the side of the object. Exemplarily, if a visibility detection result indicating that the side is invisible relative to the radar point cloud sensor is obtained, the obstacle detection system does not calculate the spatial coordinates of the center point on this side and does not consider the influence of this side when calculating the final obstacle prediction result.

[0056] In step 207, the obstacle detection system corrects the initial obstacle prediction result based on the spatial coordinates of the center point on the side to obtain the final obstacle prediction result, and the final obstacle prediction result indicates the final position of the obstacle.

[0057] Optionally, after calculating the spatial coordinates of the center point on the side, the obstacle detection system corrects the initial obstacle prediction result based on the spatial coordinates of the center point on the side to obtain the final obstacle prediction result, where the final obstacle prediction result indicates the final position of the obstacle.

[0058] Exemplarily, the obstacle detection system corrects the initial obstacle prediction result based on the spatial coordinates of the center point on the side to obtain the final obstacle prediction result, including: correcting the initial position of the obstacle based on the spatial coordinates of the center point on the side to obtain the final position of the obstacle; determining the final obstacle prediction result based on the final position of the obstacle and the shape of the obstacle.

[0059] In a possible implementation, the method for correcting the initial position of the obstacle based on the spatial coordinates of the center point on the side includes, but is not limited to: the obstacle detection system generates an intermediate position of the obstacle based on the spatial coordinates of the center point on the side, and compares the intermediate position of the obstacle with the initial position of the obstacle; if the distance between the intermediate position of the obstacle and the initial position of the obstacle is greater than the distance threshold, the intermediate position of the obstacle is used to replace the initial position of the obstacle as the final position of the obstacle; if the distance between the intermediate position of the obstacle and the initial position of the obstacle is less than or equal to the distance threshold, the midpoint of the line connecting the intermediate position of the obstacle and the initial position of the obstacle is used as the final position of the obstacle. After determining the final position of the obstacle, the obstacle detection system then determines the final obstacle prediction result based on the final position of the obstacle and the shape of the obstacle.

[0060] Optionally, after determining the final obstacle prediction result, the obstacle detection system predicts the trajectory of the obstacle based on the final obstacle prediction result and the shape of the obstacle included in the initial obstacle prediction result, and then combines the type of the obstacle, so as to adjust the driving speed and driving direction of the vehicle, reduce the collision risk between the vehicle and the obstacle, and thus ensure the driving safety of the vehicle.

[0061] In the embodiment of the present application, by collecting the information of the reflected signal corresponding to the radar point cloud sensor, the point cloud data of the obstacle is obtained, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle; then, according to the spatial coordinates of the reflection points of the obstacle, the spatial coordinates of the center point of the obstacle are calculated, which are used to generate an initial obstacle prediction result in combination with the spatial coordinates of the reflection points; based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, the visibility detection result of the side is determined; after it is determined that any side is visible relative to the radar point cloud sensor, based on the point cloud data of this side, the spatial coordinates of the center point of this side are calculated, which are used to correct the initial obstacle prediction result. By calculating the spatial coordinates of the center point of the visible surface, the position of the obstacle is corrected, improving the accuracy of the obstacle position detection and making the process of predicting the position of the obstacle more convenient to implement.

[0062] See Figure 3 , the embodiment of the present application provides a detection device for an obstacle, and the device includes:

[0063] A collection module 301, configured to collect the information of the reflected signal corresponding to the laser pulse emitted by the radar point cloud sensor;

[0064] A first generation module 302, configured to generate the point cloud data of the obstacle based on the information of the reflected signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that generates the reflected signal by reflecting the laser pulse;

[0065] A first calculation module 303, configured to calculate the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle;

[0066] A second generation module 304, configured to generate an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points, and the initial obstacle prediction result indicates the initial position and shape of the obstacle;

[0067] An acquisition module 305, configured to obtain the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, where the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located;

[0068] A second calculation module 306, configured to, in response to obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, calculate the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle;

[0069] The correction module 307 is configured to correct the initial obstacle prediction result based on the spatial coordinates of the center point of the side surface to obtain the final obstacle prediction result, and the final obstacle prediction result indicates the final position of the obstacle.

[0070] In a possible implementation, the information of the reflected signal includes the duration from emitting the laser pulse to receiving the reflected signal and the propagation direction of the reflected signal. The first generation module 302 is configured to calculate the distance between the reflection point and the radar point cloud sensor based on the duration from emitting the laser pulse to receiving the reflected signal; generate the spatial coordinates of the reflection point based on the propagation direction of the reflected signal and the distance between the reflection point and the radar point cloud sensor.

[0071] In a possible implementation, the acquisition module 305 is configured to determine the boundary points of the side surface based on the point cloud data of the side surface of the obstacle; calculate the length of the point cloud region of the side surface based on the spatial coordinates of the boundary points; count the number of points included in the side surface; and obtain the visibility detection result indicating that the side surface is visible relative to the radar point cloud sensor in response to the ratio of the top-down length of the annotation box to the length of the point cloud region of the side surface being greater than the ratio threshold and the number of points included in the side surface being greater than the number threshold.

[0072] In a possible implementation, the second generation module 304 is configured to determine the initial position of the obstacle based on the spatial coordinates of the center point of the obstacle; calculate the shape of the obstacle based on the spatial coordinates of the reflection points of the obstacle.

[0073] In a possible implementation, the correction module 307 is configured to correct the initial position of the obstacle based on the spatial coordinates of the center point of the side surface to obtain the final position of the obstacle; determine the final obstacle prediction result based on the final position of the obstacle and the shape of the obstacle.

[0074] In a possible implementation, the correction module 307 is further configured to predict the trajectory of the obstacle based on the final obstacle prediction result and the shape.

[0075] This device acquires the point cloud data of obstacles by collecting the information of the reflected signals corresponding to the radar point cloud sensor. Among them, the point cloud data includes the spatial coordinates of the reflection points of the obstacles; then, based on the spatial coordinates of the reflection points of the obstacles, the spatial coordinates of the center point of the obstacles are calculated, which are used to generate the initial obstacle prediction result in combination with the spatial coordinates of the reflection points; the visibility detection result of the side is judged based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle; after determining that any side is visible relative to the radar point cloud sensor, the spatial coordinates of the center point of this side are calculated based on the point cloud data of this side, which are used to correct the initial obstacle prediction result. By calculating the spatial coordinates of the center point of the visible surface to correct the position of the obstacle, the accuracy of obstacle position detection is improved, and the process of predicting the position of the obstacle is more convenient to implement.

[0076] It should be noted that when the device provided in the above embodiment realizes its functions, only the above-mentioned division of each functional module is used for illustration. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device provided in the above embodiment and the method embodiment belong to the same concept, and the specific implementation process can be seen in the method embodiment, which will not be repeated here.

[0077] In an exemplary embodiment, a computer-readable storage medium is further provided. At least one computer program is stored in the computer-readable storage medium, and the at least one computer program is loaded and executed by the processor of the computer device to enable the computer to implement any of the above obstacle detection methods.

[0078] In a possible implementation manner, the above computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0079] In an exemplary embodiment, a computer program product or a computer program is further provided. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to enable the computer device to implement any of the above obstacle detection methods.

[0080] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data for analysis, stored data, displayed data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with the relevant laws, regulations and standards of relevant countries and regions. For example, the information of the reflected signal, the point cloud data of the obstacle, the spatial coordinates of the center point of the obstacle, the spatial coordinates of the center point of the side, the initial obstacle prediction result and the final obstacle prediction result involved in this application are all obtained under full authorization.

[0081] It should be understood that the term "a plurality of" mentioned herein refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.

[0082] It should be noted that the terms "first", "second", etc. (if any) in the description and claims of this application are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. On the contrary, they are only examples of devices and methods consistent with some aspects of this application as detailed in the appended claims.

[0083] The above are only exemplary embodiments of this application and are not intended to limit this application. Any modifications, equivalent replacements, improvements, etc. made within the principles of this application shall be included in the protection scope of this application.

Claims

1. A method for detecting an obstacle, characterized in that The method includes: Collecting information of the reflection signal corresponding to the laser pulse emitted by the radar point cloud sensor; Generating point cloud data of the obstacle based on the information of the reflection signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflection signal; Calculating the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle; Generating an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points, where the initial obstacle prediction result indicates the initial position and shape of the obstacle; Obtaining the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, where the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located; In response to obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, calculating the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle; Correcting the initial obstacle prediction result based on the spatial coordinates of the center point of the side to obtain a final obstacle prediction result, where the final obstacle prediction result indicates the final position of the obstacle.

2. The method according to claim 1, wherein The information of the reflection signal includes the duration from emitting the laser pulse to receiving the reflection signal and the propagation direction of the reflection signal. Generating the point cloud data of the obstacle based on the information of the reflection signal includes: Calculating the distance between the reflection point and the radar point cloud sensor based on the duration from emitting the laser pulse to receiving the reflection signal; Generating the spatial coordinates of the reflection point based on the propagation direction of the reflection signal and the distance between the reflection point and the radar point cloud sensor.

3. The method according to claim 1, wherein Obtaining the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle includes: Determining the boundary points of the side based on the point cloud data of the side of the obstacle; Calculating the length of the point cloud area of the side based on the spatial coordinates of the boundary points; Counting the number of points included in the side; In response to the ratio of the top-down length of the annotation box to the length of the point cloud area of the side being greater than a ratio threshold and the number of points included in the side being greater than a number threshold, obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor.

4. The method according to claim 1, wherein Generating the initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points includes: Determining the initial position of the obstacle based on the spatial coordinates of the center point of the obstacle; Calculating the shape of the obstacle based on the spatial coordinates of the reflection points of the obstacle.

5. The method according to claim 4, wherein Correcting the initial obstacle prediction result based on the spatial coordinates of the center point of the side to obtain the final obstacle prediction result includes: Correcting the initial position of the obstacle based on the spatial coordinates of the center point of the side to obtain the final position of the obstacle; Determine the final obstacle prediction result based on the final position and the shape of the obstacle.

6. The method according to claim 1, wherein After correcting the initial obstacle prediction result based on the spatial coordinates of the center point of the side to obtain the final obstacle prediction result, it further includes: Predict the trajectory of the obstacle based on the final obstacle prediction result and the shape.

7. A detection device for obstacles, characterized in that, The device includes: An acquisition module, configured to acquire information on the reflection signal corresponding to the laser pulse emitted by the radar point cloud sensor; A first generation module, configured to generate point cloud data of the obstacle based on the information on the reflection signal, where the point cloud data includes the spatial coordinates of the reflection points of the obstacle, and the reflection point is the point that reflects the laser pulse to generate the reflection signal; A first calculation module, configured to calculate the spatial coordinates of the center point of the obstacle based on the spatial coordinates of the reflection points of the obstacle; A second generation module, configured to generate an initial obstacle prediction result based on the spatial coordinates of the center point of the obstacle and the spatial coordinates of the reflection points, where the initial obstacle prediction result indicates the initial position and shape of the obstacle; An acquisition module, configured to obtain the visibility detection result of the side based on the top-down length of the annotation box of the obstacle and the point cloud data of the side of the obstacle, where the visibility detection result indicates whether the side is visible relative to the radar point cloud sensor, and the annotation box is used to frame the area where the obstacle is located; A second calculation module, configured to, in response to obtaining the visibility detection result indicating that the side is visible relative to the radar point cloud sensor, calculate the spatial coordinates of the center point of the side based on the point cloud data of the side of the obstacle; A correction module, configured to correct the initial obstacle prediction result based on the spatial coordinates of the center point of the side to obtain the final obstacle prediction result, where the final obstacle prediction result indicates the final position of the obstacle.

8. The device according to claim 7, characterized in that, The information on the reflection signal includes the duration from emitting the laser pulse to receiving the reflection signal and the propagation direction of the reflection signal. The first generation module is configured to calculate the distance between the reflection point and the radar point cloud sensor based on the duration from emitting the laser pulse to receiving the reflection signal; and generate the spatial coordinates of the reflection point based on the propagation direction of the reflection signal and the distance between the reflection point and the radar point cloud sensor.

9. A computer program product, the computer program product includes computer instructions, and when the computer instructions are executed by a processor, the steps of the obstacle detection method according to any one of claims 1 to 6 are implemented.

10. A non-transitory computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, and the computer program is loaded and executed by a processor to implement the obstacle detection method according to any one of claims 1 to 6.

Images