Method, device, apparatus and storage medium for identifying static obstacle detection errors

Abstract

The present application relates to a method, device, equipment and storage medium for identifying static obstacle detection errors, the method comprising obtaining vehicle position information and obstacle information of each frame during the driving process of the main vehicle; determining the contour position information of each frame of the main vehicle, the contour position information of each frame of the dynamic obstacle and the contour position information of each frame of the static obstacle based on the vehicle position information and obstacle information obtained for each frame; detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame; when the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle information is erroneous. The existing problem of low efficiency and large human error in manually detecting static obstacle recognition results is solved. The present application has the advantages of improving the detection efficiency and accuracy of static obstacle recognition results.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to methods, devices, equipment and storage media for identifying static obstacle detection errors. Background Art

[0002] The autonomous driving perception module can automatically identify static obstacles on the road through a large amount of environmental information obtained by sensors and the deep learning model in the perception module. However, during the training of the deep learning model, since the training data is processed, it will be assumed that there is no problem, and it is difficult to perceive the errors in the recognition results of static obstacles when backtracking the problem. When the static obstacle is identified incorrectly, it will cause the autonomous driving vehicle to get stuck, suddenly brake, etc., thereby causing dangerous accidents such as vehicle rear-end collisions. Based on the analysis and detection of static obstacle recognition results, and using the detection results as a training guide for the deep learning model, it is beneficial to improve the recognition accuracy of the deep learning model.

[0003] The existing detection method is to manually understand the recognition results of static obstacles with the help of summarized experience knowledge to determine whether there is a static obstacle recognition error. However, there are a lot of static obstacles, and it is difficult to check the recognition results of static obstacles one by one, which is time-consuming and prone to human errors.

[0004] With respect to the above-mentioned related technologies, the inventors believe that the existing method of manually detecting the static obstacle recognition results has the defects of low efficiency and large human error. Summary of the invention

[0005] In order to improve the detection efficiency and accuracy of static obstacle recognition results, the present application provides a method, apparatus, device and storage medium for identifying static obstacle detection errors.

[0006] In a first aspect, the present application provides a method for identifying static obstacle detection errors, which has the characteristics of improving the detection efficiency and accuracy of static obstacle recognition results.

[0007] This application is achieved through the following technical solutions:

[0008] A method for identifying static obstacle detection errors comprises the following steps:

[0009] Obtain vehicle position information and obstacle information for each frame during the main vehicle's driving process;

[0010] Based on the vehicle position information and obstacle information acquired in each frame, determine the contour position information of the host vehicle in each frame, the contour position information of the dynamic obstacle in each frame, and the contour position information of the static obstacle in each frame;

[0011] Detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle have an intersection with the contour position of the static obstacle in the current frame;

[0012] When the contour position set of the host vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle has an error.

[0013] In a preferred example, the present application can be further configured as follows: the step of determining the contour position information of each frame of the main vehicle includes:

[0014] The vehicle position information of each frame obtained includes a horizontal coordinate, a vertical coordinate and a direction information;

[0015] According to the preset reference point, the initial contour position information of the main vehicle in the global coordinate system is obtained;

[0016] Based on the abscissa, ordinate and orientation information of each frame and the initial contour position information, contour position information of each frame of the host vehicle in the global coordinate system is determined.

[0017] In a preferred example, the present application may be further configured as follows: the step of determining the contour position information of each frame of the dynamic obstacle includes:

[0018] Obstacle information obtained includes position information and speed information of each frame;

[0019] When the average speed of an obstacle exceeding a preset threshold value continues to exceed a preset time, the obstacle is determined to be a dynamic obstacle, and the ID and corresponding location information of the obstacle are added to the preset dynamic obstacle set;

[0020] Based on the dynamic obstacle set, contour position information of each frame of the dynamic obstacle in a global coordinate system is determined.

[0021] In a preferred example, the present application may be further configured as follows: after the step of determining the contour position information of each frame of the dynamic obstacle, the following steps are further included:

[0022] The contour position information of each frame of the host vehicle is added to the dynamic obstacle set, and corresponds one-to-one with the contour position information of each frame of the dynamic obstacle, so as to obtain a preprocessed dynamic obstacle set.

[0023] In a preferred example, the present application may be further configured as follows: the step of determining the contour position information of each frame of the static obstacle includes:

[0024] Obstacle information obtained includes position information and speed information of each frame;

[0025] When the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than the preset distance, the obstacle is judged to be a static obstacle, and the ID and corresponding position information of the obstacle are added to the preset static obstacle set;

[0026] Based on the static obstacle set, contour position information of each frame of the static obstacle in a global coordinate system is determined.

[0027] In a preferred example, the present application may be further configured as follows: the step of detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle have an intersection with the contour position of the static obstacle in the current frame includes:

[0028] Preset time window threshold;

[0029] Obtaining contour position information of all dynamic obstacles corresponding to the time window threshold in contour position information of each frame of dynamic obstacles;

[0030] It is determined whether a set formed by the contour position information of the dynamic obstacle has an intersection with the acquired contour position information of the static obstacle.

[0031] In a preferred example, the present application may be further configured as follows: while the step of obtaining the contour position information of all dynamic obstacles corresponding to the time window threshold in the contour position information of each frame of the dynamic obstacle, the following steps are also included:

[0032] Obtaining contour position information of all main vehicles corresponding to the time window threshold in the contour position information of each frame of the main vehicle;

[0033] Determine a reference contour position information set based on the acquired contour position information of the host vehicle and the contour position information of the dynamic obstacle;

[0034] It is determined whether the reference contour position information set has an intersection with the acquired contour position information of the static obstacle.

[0035] In a preferred example, the present application may be further configured as follows: the step of determining whether the reference contour position information set and the acquired contour position information of the static obstacle have an intersection includes:

[0036] In the same coordinate system, marking the contour corresponding to the reference contour position information set and the contour of the static obstacle;

[0037] When the contour corresponding to the reference contour position information set overlaps with the contour of the static obstacle, that is, the reference contour position information set and the acquired contour position information of the static obstacle have an intersection.

[0038] In a preferred example, the present application may be further configured as follows: after the step of marking the contour corresponding to the reference contour position information set and the contour of the static obstacle, the following steps are further included:

[0039] Calculating the contour area corresponding to the reference contour position information set and the contour area of ​​the static obstacle, and the intersection area of ​​the contour corresponding to the reference contour position information set and the contour of the static obstacle;

[0040] Taking the smaller contour area as a target value, solving for the ratio of the intersection area to the target value;

[0041] When the ratio is greater than a preset ratio threshold, the reference contour position information set and the acquired contour position information of the static obstacle have an intersection.

[0042] On the second aspect, the present application provides a device for identifying static obstacle detection errors, which has the characteristics of improving the detection efficiency and accuracy of static obstacle recognition results.

[0043] This application is achieved through the following technical solutions:

[0044] A device for identifying static obstacle detection errors, comprising:

[0045] A data acquisition module is used to obtain the vehicle position information and obstacle information of each frame during the driving process of the main vehicle;

[0046] A calculation module, for determining the contour position information of the host vehicle in each frame, the contour position information of the dynamic obstacle in each frame, and the contour position information of the static obstacle in each frame based on the vehicle position information and obstacle information acquired in each frame;

[0047] The detection module is used to detect whether the outline position set of the main vehicle and the outline position set of the dynamic obstacle have an intersection with the outline position of the static obstacle in the current frame; when the outline position set of the main vehicle and the outline position set of the dynamic obstacle have an intersection with the outline position of the static obstacle, the corresponding static obstacle information has an error.

[0048] On the third aspect, the present application provides a computer device having the characteristics of improving the detection efficiency and accuracy of static obstacle recognition results.

[0049] This application is achieved through the following technical solutions:

[0050] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for identifying static obstacle detection errors when executing the computer program.

[0051] In a fourth aspect, the present application provides a computer-readable storage medium having the characteristics of improving the detection efficiency and accuracy of static obstacle recognition results.

[0052] This application is achieved through the following technical solutions:

[0053] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for identifying static obstacle detection errors are implemented.

[0054] In a fifth aspect, the present application provides a computer program product having the characteristics of improving the detection efficiency and accuracy of static obstacle recognition results.

[0055] This application is achieved through the following technical solutions:

[0056] A computer program product comprises a computer program, wherein when the computer program is executed by a processor, the steps of the above-mentioned method for identifying static obstacle detection errors are implemented.

[0057] In summary, the present application includes at least one of the following beneficial technical effects:

[0058] 1. Obtain the vehicle position information and obstacle information of each frame during the driving process of the main vehicle to determine the contour position information of each frame of the main vehicle, the contour position information of each frame of the dynamic obstacle, and the contour position information of each frame of the static obstacle. Then, based on the contour information, detect whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame. If there is an intersection, a collision occurs, and then judge that the detection result of the "static obstacle" perceived by the perception module is incorrect. With the help of the motion trajectory formed by the time difference of the contour information, automatic judgment is made, and more comprehensive data information is obtained from the time dimension. Then, the phenomenon that violates the facts is found out to infer the situation of static obstacle detection error, which makes up for the fact that the perception module is difficult to think of the inaccurate training data provided by the upstream module, and can quickly and automatically identify the situation of static obstacle detection error without manual investigation, thereby improving the detection efficiency and accuracy of static obstacle recognition results;

[0059] 2. When the average speed of an obstacle exceeding the preset threshold continues to exceed the preset time, the obstacle is judged as a dynamic obstacle to remove obstacles with relatively low speed and improve the accuracy of the judgment result;

[0060] 3. When the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than the preset distance, the obstacle is judged to be a static obstacle to improve the accuracy of the judgment result. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a flowchart of a method for identifying static obstacle detection errors according to one embodiment of the present application.

[0062] Figure 2 It is a flow chart of determining the contour position information of the main vehicle in each frame.

[0063] Figure 3 It is a schematic diagram of the relative position of GPS and the outline of the autonomous driving vehicle.

[0064] Figure 4 It is a flowchart for determining the contour position information of each frame of a dynamic obstacle.

[0065] Figure 5 It is a flowchart for determining the contour position information of each frame of a static obstacle.

[0066] Figure 6 It is a flowchart for detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle.

[0067] Figure 7 This is a structural block diagram of a device for identifying static obstacle detection errors in one embodiment of the present application. DETAILED DESCRIPTION

[0068] This specific embodiment is merely an explanation of the present application and is not a limitation of the present application. After reading this specification, those skilled in the art may make modifications to the present embodiment without any creative contribution as needed, but such modifications are protected by the patent law as long as they are within the scope of the claims of the present application.

[0069] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0070] In addition, the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article, unless otherwise specified, generally means that the associated objects before and after are in an "or" relationship.

[0071] When an autonomous driving vehicle encounters an abnormality, such as sudden braking, stopping in the middle of the road, taking over, etc., which causes a collision, the present application automatically uploads relevant data from a period of time before the collision (such as half a minute to one minute) to the moment of the collision to the cloud, so that the cloud automatically analyzes the acquired data to quickly and automatically identify situations where the perception module incorrectly detects static obstacles.

[0072] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0073] Reference Figure 1 , an embodiment of the present application provides a method for identifying static obstacle detection errors, and the main steps of the method are described as follows.

[0074] S1: Obtain vehicle position information and obstacle information for each frame during the main vehicle's driving process;

[0075] S2: Based on the acquired vehicle position information and obstacle information of each frame, determine the contour position information of each frame of the host vehicle, the contour position information of each frame of the dynamic obstacle, and the contour position information of each frame of the static obstacle;

[0076] S3: Detect whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame;

[0077] S4: When the contour position set of the host vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle has an error.

[0078] Further, S5: output corresponding static obstacle information.

[0079] Further, refer to Figure 2 , S2: The step of determining the contour position information of each frame of the main vehicle comprises:

[0080] S10: The vehicle position information of each frame obtained includes a horizontal coordinate, a vertical coordinate and a direction information;

[0081] S11: Obtaining the initial contour position information of the main vehicle in the global coordinate system according to the preset GPS position information relative to the main vehicle;

[0082] S12: Based on the horizontal coordinate, vertical coordinate and orientation information of each frame, and the initial contour position information, determine the contour position information of each frame of the main vehicle in the global coordinate system.

[0083] Specifically, when an autonomous vehicle is driving, each module will record and store the input and output data of the operation. For example, data acquisition modules such as GPS, IMU, camera, and lidar will record the data collected for each frame. At the same time, the vehicle positioning module obtains the current vehicle location information based on the information obtained from GPS and IMU.

[0084] The GPS data and IMU data are acquired to obtain the vehicle position information of each frame. The vehicle position information includes the horizontal coordinate, the vertical coordinate and the orientation information.

[0085] Furthermore, refer to Figure 3 Before the autonomous driving vehicle leaves the factory, the relative position information between GPS and the contour of the autonomous driving vehicle can be known in advance, and the initial contour position information of the main vehicle in the global coordinate system is obtained according to the preset reference point. In this embodiment, based on the position information of GPS relative to the main vehicle, the position of GPS is used as the reference point, and the left distance from the reference point to the autonomous driving vehicle is read as d4, the right distance from the reference point to the autonomous driving vehicle is d3, the front distance from the reference point to the autonomous driving vehicle is d2, and the rear distance from the reference point to the autonomous driving vehicle is d1. Therefore, the coordinate set vec formed by the four vertices A, B, C, and D corresponding to the initial contour E0 of the autonomous driving vehicle is obtained, and vec = {(-d3, d1), (d4, d1), (-d3, -d2), (-d4, -d2)}.

[0086] Assuming that the orientation information of the autonomous driving vehicle is angle, the current position of the autonomous driving vehicle is pos, i.e., the horizontal coordinate and the vertical coordinate. Based on the horizontal coordinate, vertical coordinate and orientation information of each frame, and the initial contour position information, the contour position information E of the main vehicle in each frame in the global coordinate system is determined. The specific calculation formula can be:

[0087] cos,sin=np.cos(angle),np.sin(angle);

[0088] rot_matrix=np.array([[cos,-sin],[sin,cos]]);

[0089] out=np.dot(vec,np.transpose(rot_matrix));

[0090] E=pos+out.

[0091] Further, refer to Figure 4 , S2: The step of determining the contour position information of each frame of the dynamic obstacle comprises:

[0092] S20: Obtained obstacle information includes position information and speed information of each frame;

[0093] S21: When the average speed of an obstacle exceeding a preset threshold value continues to exceed a preset time, the obstacle is determined to be a dynamic obstacle, and the ID and corresponding position information of the obstacle are added to a preset dynamic obstacle set;

[0094] S22: Based on the dynamic obstacle set, determine the contour position information of each frame of the dynamic obstacle in the global coordinate system.

[0095] Specifically, with the help of the data collected by the data collection module and the current vehicle location information obtained by the vehicle positioning module, the perception module calculates the surrounding obstacle information based on the neural network model. The obstacle information includes obstacle type information, obstacle location information, obstacle outline information, obstacle speed information, etc.

[0096] Obstacle information is obtained, including the position information and speed information of each frame of the obstacle, and when the average speed of the obstacle exceeds the preset threshold and lasts for more than the preset time, the obstacle is judged as a dynamic obstacle to remove obstacles with relatively small speeds and improve the accuracy of the judgment result. In this embodiment, when the average speed of the obstacle exceeds 0.5m / s and the duration exceeds 3s, it is judged as a dynamic obstacle and a unique obstacle ID is assigned.

[0097] Create a dynamic obstacle set dynamic_obstacles={}, whose initial state is empty, and add the ID and corresponding position information of each dynamic obstacle sensed to the dynamic obstacle set in sequence.

[0098] Based on the dynamic obstacle set, the contour position information D of each frame of the dynamic obstacle in the global coordinate system is determined. Specifically, the position information of the dynamic obstacles in the dynamic obstacle set is input into the preset neural network model to output the contour position information of each frame of the dynamic obstacle, and then combined with the coordinate conversion formula, the contour position information D of each frame of the dynamic obstacle in the global coordinate system is obtained. Among them, the preset neural network model is obtained by obtaining vehicle collection data, manually annotating the obstacle 3D frame, and inputting it into the deep learning model for training.

[0099] Further, S23: the contour position information of each frame of the main vehicle is added to the dynamic obstacle set, that is, the contour information E of the main vehicle is added to dynamic_obstacles, and the contour position information E of each frame of the main vehicle corresponds to the contour position information D of each frame of the dynamic obstacle one by one, which is equivalent to treating the contour of the main vehicle as a dynamic obstacle contour, which is used to subsequently determine whether it overlaps with the static obstacle, and then obtain the pre-processed dynamic obstacle set, that is, the effective dynamic obstacle set, as shown in Table 1.

[0100] Table 1

[0101] time Dynamic obstacle contours T0 E0, D0, D1… T1 E1, D0, D1, D2… … … Tn …

[0102] Further, refer to Figure 5 , S2: The step of determining the contour position information of each frame of the static obstacle comprises:

[0103] S30: Obtained obstacle information includes position information and speed information of each frame;

[0104] S31: When the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than the preset distance, the obstacle is determined to be a static obstacle, and the ID and location information corresponding to the obstacle are added to the preset static obstacle set;

[0105] S32: Based on the static obstacle set, determine the contour position information of each frame of the static obstacle in the global coordinate system.

[0106] Specifically, obstacle information is obtained, including the position information and speed information of each frame of the obstacle, and when the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than a preset distance, the obstacle is judged to be a static obstacle to improve the accuracy of the judgment result. In this embodiment, when the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than 50m, it is judged to be a static obstacle and a unique obstacle ID is assigned.

[0107] Create a static obstacle set static_obstacles={}, whose initial state is empty, and add the ID and location information of each static obstacle sensed to the static obstacle set in sequence.

[0108] Based on the static obstacle set, the contour position information S of each frame of the static obstacle in the global coordinate system is determined. Specifically, the position information of the static obstacles in the static obstacle set is input into the above-mentioned preset neural network model to output the contour position information of each frame of the static obstacle, and then combined with the coordinate conversion formula, the contour position information S of each frame of the static obstacle in the global coordinate system is obtained. Then, a valid static obstacle set is obtained, as shown in Table 2.

[0109] Table 2

[0110] time Static obstacle outlines T0 S0,S1… T1 S0, S1, S2… … … Tn …

[0111] Further, refer to Figure 6 S3: The step of detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame includes:

[0112] S030: preset time window threshold;

[0113] S031: respectively obtaining the contour position information of all the main vehicles and the contour position information of all the dynamic obstacles corresponding to the time window threshold in the contour position information of each frame of the main vehicle and the contour position information of each frame of the dynamic obstacles;

[0114] S032: Determine a reference contour position information set based on the acquired contour position information of the host vehicle and the contour position information of the dynamic obstacle;

[0115] S033: Calculate the contour area corresponding to the reference contour position information set and the contour area corresponding to the contour position information of the static obstacle, as well as the intersection area of ​​the reference contour position information set and the contour position information of the static obstacle obtained in the current frame;

[0116] S034: Taking the small contour area as the target value, solving the ratio of the intersection area to the target value; when the ratio is greater than a preset ratio threshold, the contour position set of the main vehicle and the contour position set of the dynamic obstacle have an intersection with the contour position of the static obstacle in the current frame.

[0117] Specifically, create an error static obstacle set falas_alarm_obstacle={}, and initialize falas_alarm_obstacles to an empty set.

[0118] At the same time, a time window threshold t is preset. In this embodiment, the time window threshold t may be 15 seconds, because a static obstacle recognition misjudgment occurring 15 seconds ago will not have a particularly large impact on the autonomous driving vehicle.

[0119] For the static obstacle contour S corresponding to the current frame T in the valid static obstacle set, all valid dynamic obstacle contours corresponding to the time range (Tt, T+t) are obtained from the valid dynamic obstacle set, that is, the contour position information of all the main vehicles and the contour position information of all dynamic obstacles corresponding to the time window threshold in the contour position information of each frame of the main vehicle and the contour position information of each frame of the dynamic obstacles are obtained respectively, so as to obtain the motion trajectory information of the main vehicle and the motion trajectory information of the dynamic obstacles with the help of the time difference of the contour information, and determine whether they overlap with the static obstacles.

[0120] Based on the acquired contour position information of the host vehicle and the contour position information of the dynamic obstacle, a reference contour position information set is determined.

[0121] Calculate the contour area corresponding to the reference contour position information set and the contour area corresponding to the contour position information of the static obstacle, that is, calculate the contour area enclosed by the reference contour position information set in the global coordinate system and the contour area enclosed by the contour position information of the static obstacle in the current frame in the global coordinate system.

[0122] Based on the contour area enclosed by the reference contour position information set in the global coordinate system and the contour area enclosed by the contour position information of the static obstacle in the current frame in the global coordinate system, the intersection area of ​​the reference contour position information set and the obtained contour position information of the static obstacle in the current frame is calculated, that is, the contour enclosed by the reference contour position information set in the global coordinate system and the contour enclosed by the contour position information of the static obstacle in the current frame in the global coordinate system are marked in the global coordinate system. If there is an intersection, it is assumed that there is an intersection, and a collision will occur if there is an intersection. At this time, the overlapping area intersection_area of ​​the two contours is calculated.

[0123] Determine the size of the contour area enclosed by the reference contour position information set in the global coordinate system and the contour area enclosed by the contour position information of the static obstacle in the current frame in the global coordinate system, and use the smaller contour area as the target value to solve the ratio of the overlapping area intersection_area to the target value.

[0124] When the ratio is greater than a preset ratio threshold, the contour position set of the host vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame, that is, the host vehicle has collided. In this embodiment, the ratio threshold may be 0.3.

[0125] At this time, the corresponding static obstacle of the current frame has an error, so that the outline S of the static obstacle is added to the error static obstacle set falas_alarm_obstacle, and the static obstacle information is output to visualize the recognized error static obstacle.

[0126] Furthermore, it is manually checked whether the static obstacles in the error static obstacle set falas_alarm_obstacle collide with the main vehicle, and it is confirmed again whether it is a static obstacle that is misidentified.

[0127] Furthermore, a method for identifying static obstacle detection errors obtains vehicle position information and obstacle information of each frame during the driving process of the main vehicle to determine the contour position information of each frame of the main vehicle, the contour position information of each frame of the dynamic obstacle, and the contour position information of each frame of the static obstacle. Based on the contour information, it is detected whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame. If there is an intersection, a collision occurs, and then it can be judged that the detection result of the "static obstacle" perceived by the perception module is incorrect. The trajectory formed by the time difference of the contour information is used for judgment, and more comprehensive data information is obtained from the time dimension. Then, phenomena that violate the facts are found out to infer the situation of static obstacle detection errors, which makes up for the fact that the perception module is difficult to think of the inaccurate training data provided by the upstream module, and can quickly and automatically identify the situation of static obstacle detection errors without manual investigation one by one, thereby improving the detection efficiency and accuracy of the static obstacle recognition results.

[0128] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0129] Reference Figure 7 The embodiment of the present application further provides a device for identifying a static obstacle detection error, and the device for identifying a static obstacle detection error corresponds one-to-one with the method for identifying a static obstacle detection error in the above embodiment. The device for identifying a static obstacle detection error includes:

[0130] A data acquisition module is used to obtain the vehicle position information and obstacle information of each frame during the driving process of the main vehicle;

[0131] A calculation module, for determining the contour position information of the host vehicle in each frame, the contour position information of the dynamic obstacle in each frame, and the contour position information of the static obstacle in each frame based on the vehicle position information and obstacle information acquired in each frame;

[0132] The detection module is used to detect whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle; when the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle information is wrong.

[0133] A device for identifying static obstacle detection errors also includes:

[0134] The display module is used to visually display erroneous static obstacle information.

[0135] For the specific definition of a device for identifying static obstacle detection errors, please refer to the definition of a method for identifying static obstacle detection errors above, which will not be repeated here. Each module in the above-mentioned device for identifying static obstacle detection errors can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0136] In one embodiment, a computer device is provided, which may be a server. The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for identifying static obstacle detection errors is implemented.

[0137] In one embodiment, a computer-readable storage medium is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0138] S1: Obtain vehicle position information and obstacle information for each frame during the main vehicle's driving process;

[0139] S2: Based on the acquired vehicle position information and obstacle information of each frame, determine the contour position information of each frame of the host vehicle, the contour position information of each frame of the dynamic obstacle, and the contour position information of each frame of the static obstacle;

[0140] S3: Detect whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle in the current frame;

[0141] S4: When the contour position set of the host vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle information is wrong.

[0142] In one embodiment, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, a method for identifying static obstacle detection errors is implemented.

[0143] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0144] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

Claims

1. A method for identifying static obstacle detection errors, characterized in that: The following steps are involved: Obtain vehicle position information and obstacle information for each frame during the main vehicle's driving process; Based on the vehicle position information and obstacle information acquired in each frame, determine the contour position information of the host vehicle in each frame, the contour position information of the dynamic obstacle in each frame, and the contour position information of the static obstacle in each frame; Detecting whether the contour position set of the main vehicle and the contour position set of the dynamic obstacle have an intersection with the contour position of the static obstacle in the current frame, including: Preset time window threshold; Obtaining contour position information of all dynamic obstacles corresponding to the time window threshold in contour position information of each frame of dynamic obstacles; Obtaining contour position information of all main vehicles corresponding to the time window threshold in the contour position information of each frame of the main vehicle; Add the contour position information of each frame of the main vehicle to the dynamic obstacle set; Determine a reference contour position information set based on the acquired contour position information of the host vehicle and the contour position information of the dynamic obstacle; Determine whether the reference contour position information set and the acquired contour position information of the static obstacle have an intersection; When the contour position set of the host vehicle and the contour position set of the dynamic obstacle intersect with the contour position of the static obstacle, the corresponding static obstacle has an error.

2. A method for identifying static obstacle detection errors according to claim 1, characterized in that: The step of determining the contour position information of each frame of the main vehicle comprises: The vehicle position information of each frame obtained includes a horizontal coordinate, a vertical coordinate and a direction information; According to the preset reference point, the initial contour position information of the main vehicle in the global coordinate system is obtained; Based on the abscissa, ordinate and orientation information of each frame and the initial contour position information, contour position information of each frame of the host vehicle in the global coordinate system is determined.

3. The method for identifying static obstacle detection errors according to claim 1, characterized in that: The step of determining the contour position information of each frame of the dynamic obstacle comprises: Obstacle information obtained includes position information and speed information of each frame; When the average speed of an obstacle exceeding a preset threshold value continues to exceed a preset time, the obstacle is determined to be a dynamic obstacle, and the ID and corresponding location information of the obstacle are added to the preset dynamic obstacle set; Based on the dynamic obstacle set, contour position information of each frame of the dynamic obstacle in a global coordinate system is determined.

4. A method for identifying static obstacle detection errors according to claim 3, characterized in that: After the step of determining the contour position information of each frame of the dynamic obstacle, the following steps are also included: The contour position information of each frame of the host vehicle is added to the dynamic obstacle set, and corresponds one-to-one with the contour position information of each frame of the dynamic obstacle, so as to obtain a preprocessed dynamic obstacle set.

5. The method for identifying static obstacle detection errors according to claim 1, characterized in that: The step of determining the contour position information of each frame of the static obstacle comprises: Obstacle information obtained includes position information and speed information of each frame; When the speed of the obstacle in the current frame is 0 and the relative distance between the obstacle and the main vehicle is less than the preset distance, the obstacle is judged to be a static obstacle, and the ID and corresponding position information of the obstacle are added to the preset static obstacle set; Based on the static obstacle set, contour position information of each frame of the static obstacle in a global coordinate system is determined.

6. The method for identifying static obstacle detection errors according to claim 1, characterized in that: The step of determining whether the reference contour position information set and the acquired contour position information of the static obstacle have an intersection includes: In the same coordinate system, marking the contour corresponding to the reference contour position information set and the contour of the static obstacle; When the contour corresponding to the reference contour position information set overlaps with the contour of the static obstacle, that is, the reference contour position information set and the acquired contour position information of the static obstacle have an intersection.

7. A method for identifying static obstacle detection errors according to claim 6, characterized in that: After the step of marking the contour corresponding to the reference contour position information set and the contour of the static obstacle, the following steps are also included: Calculating the contour area corresponding to the reference contour position information set and the contour area of ​​the static obstacle, and the intersection area of ​​the contour corresponding to the reference contour position information set and the contour of the static obstacle; Taking the smaller contour area as a target value, solving for the ratio of the intersection area to the target value; When the ratio is greater than a preset ratio threshold, the reference contour position information set and the acquired contour position information of the static obstacle have an intersection.

8. A device for identifying static obstacle detection errors, characterized in that: include: A data acquisition module is used to obtain the vehicle position information and obstacle information of each frame during the driving process of the main vehicle; A calculation module, for determining the contour position information of the host vehicle in each frame, the contour position information of the dynamic obstacle in each frame, and the contour position information of the static obstacle in each frame based on the vehicle position information and obstacle information acquired in each frame; A detection module, used to detect whether the outline position set of the main vehicle and the outline position set of the dynamic obstacle have an intersection with the outline position of the static obstacle in the current frame, including a preset time window threshold; Obtaining the contour position information of all dynamic obstacles corresponding to the time window threshold in the contour position information of each frame of the dynamic obstacles; obtaining the contour position information of all host vehicles corresponding to the time window threshold in the contour position information of each frame of the host vehicle; and adding the contour position information of each frame of the host vehicle to the dynamic obstacle set; Based on the acquired contour position information of the main vehicle and the contour position information of the dynamic obstacle, a reference contour position information set is determined; it is judged whether the reference contour position information set has an intersection with the acquired contour position information of the static obstacle; when the contour position set of the main vehicle and the contour position set of the dynamic obstacle have an intersection with the contour position of the static obstacle, the corresponding static obstacle information has an error.

9. A computer device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product, characterized in that The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 7 when being executed by a processor.

Images