Vehicle obstacle processing method and device, medium and equipment
By performing obstacle detection and position information merging processing on multiple image frames, the problem of low accuracy in static obstacle tracking in the prior art is solved, and more efficient obstacle tracking and repair are achieved.
Patent Information
- Application Number
- CN202311808817.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-26
- Publication Date
- 2025-06-27
AI Technical Summary
The existing 3D obstacle detection model based on point clouds has limited tracking capabilities when a large number of static obstacles exist, which is prone to tracking loss, resulting in low tracking accuracy of static obstacles.
By acquiring multiple image frames, obstruction detection is performed, the position information of the static obstacle is determined, and these position information are combined to obtain the position information of the target obstacle. The specific methods include performing initial detection based on the obstacle detection model, extracting static obstacles, and improving tracking accuracy through the merging of position information.
It improves the tracking accuracy of static obstacles, effectively avoids tracking loss, improves computing efficiency, and realizes tracking and repair of static obstacles.
Smart Images
Figure CN120220102A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of automotive technologies, and in particular, to a method, apparatus, medium, and device for processing vehicle obstacles. Background Art
[0002] In recent years, with the continuous development of automotive technologies and the improvement of economic levels, automobiles, as major consumer goods, have entered thousands of households. When purchasing an automobile, users not only care about the price, appearance, and performance of the automobile, but also the safety of the automobile, which is a major factor for users to consider when making a purchase.
[0003] Obstacle detection during the driving of intelligent vehicles is an important part of automotive safety. Currently, the tracking ability of 3D obstacle detection models based on point clouds is limited. Especially in the presence of a large number of static obstacles, it is very easy to lose tracking. Therefore, the prior art has the technical problem of low tracking accuracy for static obstacles. Summary of the Invention
[0004] To solve the above technical problems, the present disclosure provides a method, apparatus, medium, and device for processing vehicle obstacles to improve the tracking accuracy of static obstacles.
[0005] The present disclosure provides a method for processing vehicle obstacles, including:
[0006] Obtaining a plurality of image frames collected for vehicle obstacles;
[0007] Performing obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, where each of the image frames includes at least one of the static obstacles;
[0008] Performing a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles.
[0009] In some embodiments, performing obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles includes:
[0010] Performing obstacle detection on the plurality of image frames based on an obstacle detection model to determine the position information of a plurality of initial obstacles;
[0011] Extracting a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles, and determining the position information of the plurality of static obstacles.
[0012] In some embodiments, extracting a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles includes at least one of the following:
[0013] Determine the initial obstacles whose position information is detected in only one image frame among the multiple initial obstacles as static obstacles;
[0014] Based on the position information of the multiple initial obstacles, determine the position change distance of each initial obstacle between adjacent image frames, and determine the initial obstacles with a position change distance less than the first distance threshold as static obstacles.
[0015] In some embodiments, the determining the position information of the multiple static obstacles includes:
[0016] For each of the static obstacles, determine the average value of the position information of at least one initial obstacle corresponding to the static obstacle as the position information of the static obstacle.
[0017] In some embodiments, performing a merging process on the position information of the multiple static obstacles to obtain the position information of multiple target obstacles includes:
[0018] For each of the static obstacles, determine the static obstacles with a distance less than the second distance threshold from the static obstacle as the obstacles to be merged, and perform a merging process on the position information of the static obstacle and the obstacles to be merged to obtain the position information of a target obstacle, and further obtain the position information of multiple target obstacles.
[0019] In some embodiments, for each of the static obstacles, determine the static obstacles with a distance less than the second distance threshold from the static obstacle as the obstacles to be merged, and perform a merging process on the position information of the static obstacle and the obstacles to be merged to obtain the position information of a corresponding target obstacle, including:
[0020] Put the multiple static obstacles into a first set;
[0021] Determine each of the static obstacles in the first set as an obstacle to be processed, and move the obstacle to be processed from the first set to a second set;
[0022] Determine the static obstacles with a distance less than the second distance threshold from the obstacle to be processed in the first set as the obstacles to be merged, and move the obstacles to be merged from the first set to the second set;
[0023] Perform a merging process on the position information of the obstacle to be processed and the obstacles to be merged in the second set to obtain the position information of a target obstacle.
[0024] In some embodiments, before performing a merging process on the position information of the multiple static obstacles to obtain the position information of multiple target obstacles, the method further includes:
[0025] Sort the position information of the multiple static obstacles in descending order according to the number of corresponding position information.
[0026] The present disclosure also provides a vehicle obstacle processing device, including:
[0027] An acquisition module, configured to acquire a plurality of image frames collected for vehicle obstacles;
[0028] A determination module, configured to perform obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, wherein each of the image frames includes at least one of the static obstacles;
[0029] A merging module, configured to perform a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles.
[0030] The present disclosure also provides a computer-readable storage medium storing a program or instructions, and the program or instructions cause a computer to execute the steps of any one of the above methods.
[0031] The present disclosure also provides an electronic device, including:
[0032] One or more processors;
[0033] A memory, configured to store one or more programs or instructions;
[0034] The processor is configured to execute the steps of any one of the above methods by calling the programs or instructions stored in the memory.
[0035] The technical solution provided by the embodiments of the present disclosure has the following advantages compared with the prior art:
[0036] The technical solution provided by the embodiments of the present disclosure first acquires a plurality of image frames collected for vehicle obstacles, then performs obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, each image frame includes at least one static obstacle, and finally performs a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles. The embodiments of the present disclosure determine the position information of a plurality of static obstacles by performing obstacle detection on a plurality of image frames, and perform a merging process based on the position information of the static obstacles, which has high computing efficiency and can effectively avoid the occurrence of tracking loss, realizes the tracking and repair of static obstacles, and solves the technical problem of low tracking accuracy of static obstacles. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure.
[0038] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0039] Figure 1 It is a flowchart of a method for processing vehicle obstacles provided by an embodiment of the present disclosure;
[0040] Figure 2 It is a flowchart of step S120 in an embodiment of the present disclosure;
[0041] Figure 3 It is a flowchart of step S130 in an embodiment of the present disclosure;
[0042] Figure 4 It is a structural block diagram of a device for processing vehicle obstacles provided by an embodiment of the present disclosure;
[0043] Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure. Detailed implementation manners
[0044] In order to be able to more clearly understand the above objects, features, and advantages of the present disclosure, the following will further describe the solutions of the present disclosure. It should be noted that, without conflict, the embodiments of the present disclosure and the features in the embodiments can be combined with each other.
[0045] Many specific details are set forth in the following description in order to fully understand the present disclosure, but the present disclosure can also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only a part of the embodiments of the present disclosure, rather than all of the embodiments.
[0046] Figure 1 It is a flowchart of a method for processing vehicle obstacles provided by an embodiment of the present disclosure. This method is applicable to the scenario of vehicle intelligent driving, can effectively track static obstacles, and can be applied to autonomous driving vehicles and manned vehicles. This method can be executed by a device for processing vehicle obstacles, and the device for processing vehicle obstacles can be implemented in a software and / or hardware manner. As Figure 1 shown, the method includes the following steps:
[0047] S110. Obtain a plurality of image frames collected for vehicle obstacles.
[0048] S120. Detect obstacles in multiple image frames to determine the position information of multiple static obstacles.
[0049] Among them, each image frame includes at least one static obstacle.
[0050] S130. Perform a merging process on the position information of multiple static obstacles to obtain the position information of multiple target obstacles.
[0051] The method for processing vehicle obstacles provided by the embodiments of the present disclosure detects obstacles in multiple image frames to determine the position information of multiple static obstacles, and performs a merging process based on the position information of the static obstacles, which has high computational efficiency, can effectively avoid the occurrence of tracking loss, realizes the tracking and repair of static obstacles, and solves the technical problem of low tracking accuracy of static obstacles.
[0052] In some embodiments, step S110 specifically includes:
[0053] Collect multiple image frames of vehicle obstacles using in-vehicle sensors. Among them, the in-vehicle sensors can be cameras, lidars, etc., and then the original information collected by the in-vehicle sensors can be used to obtain data containing the environmental information of the vehicle; the number of sensors can be one or multiple. For example, a camera or lidar can be set in front of the vehicle, and then the image or point cloud information in front of the vehicle can be collected by the sensor, and then data of the environmental information in front of the vehicle can be generated. Or multiple fisheye cameras can be set at multiple different positions of the vehicle, and then data containing the environmental information around the vehicle can be generated using the multiple collected images. It should be noted that point cloud data is required in the subsequent steps, and the process of obtaining point cloud data from the original information collected by the sensor can be achieved using existing methods, such as obtaining the depth information of each position in the image to obtain point cloud data, etc., which is not limited here.
[0054] In some embodiments, as Figure 2 shown, step S120 specifically includes:
[0055] S1201. Detect obstacles in multiple image frames based on an obstacle detection model to determine the position information of multiple initial obstacles.
[0056] In some embodiments, the following steps can be performed on each image frame detected by the in-vehicle sensor within a preset time:
[0057] Among them, the preset time can refer to taking 15 seconds as a group (clip) of image frames, and the acquisition frequency of point cloud data is 10Hz. Therefore, a group of image frames includes 150 image frames.
[0058] First, obtain the IMU information of the ego vehicle in the current image frame. The Inertial Measurement Unit (IMU) is a device that measures the three-axis attitude angles and accelerations of an object, usually including a gyroscope, accelerometers, and sometimes also a magnetometer. The gyroscope is used to measure the angular velocities of the three axes, the accelerometers are used to measure the accelerations of the three axes, and the magnetometer provides orientation information. In intelligent driving, the IMU can provide vehicle position and attitude information when other sensors are unavailable.
[0059] Then, obtain the position information of the obstacles in the current image frame relative to the ego vehicle.
[0060] The embodiments of the present disclosure mainly perform obstacle detection on vehicles on the road. In other embodiments, obstacle detection can also be performed on other objects. According to the point cloud data, the position information of the obstacles relative to the ego vehicle can be obtained through an obstacle detection model, including the translation and rotation information of the obstacles relative to the ego vehicle. The obstacle detection model adopted in the embodiments of the present disclosure can be a 3D obstacle detection model.
[0061] Finally, based on the IMU information of the ego vehicle and the position information of the obstacles relative to the ego vehicle, calculate the position of the obstacles in the current image frame in the world coordinate system. By integrating the IMU information of the ego vehicle and the position information of the obstacles relative to the ego vehicle, the position of each obstacle in the world coordinate system can be calculated. The vehicle head unit can determine the point cloud data corresponding to the obstacles in the current image frame from the data collected by the vehicle-mounted sensors. In addition, since the position of the vehicle-mounted sensors on the vehicle is fixed, the specific extrinsic parameters of the vehicle-mounted sensors can be determined, and then the specific position of the vehicle-mounted sensors in the vehicle coordinate system can be obtained using existing methods. The intrinsic parameters of the vehicle-mounted sensors are already known. Moreover, the position of the point cloud corresponding to the original information collected by the vehicle-mounted sensors in the vehicle-mounted sensor coordinate system can be obtained during the collection process. With the intrinsic and extrinsic parameters of the vehicle-mounted sensors determined, the point cloud data of the obstacles in the current image frame in the vehicle coordinate system can be obtained using existing coordinate transformation methods. In addition, the transformation relationship between the vehicle coordinate system and the world coordinate system can be obtained through existing methods such as vehicle positioning. At this time, using existing coordinate transformation methods, the coordinates of each point cloud in the point cloud data of the obstacles in the world coordinate system can be determined, that is, the position of the obstacles in the current image frame in the world coordinate system can be obtained.
[0062] For 150 image frames within a preset time, calculate the world coordinates of each obstacle in each image frame through the above steps, and add a tracking number (track_id) to each obstacle.
[0063] S1202. Extract multiple static obstacles from the multiple initial obstacles based on the position information of the multiple initial obstacles, and determine the position information of the multiple static obstacles.
[0064] In some embodiments, extracting multiple static obstacles from the multiple initial obstacles may include at least one of the following two cases:
[0065] Case 1. Determine the position change distance of each initial obstacle between adjacent image frames based on the position information of the multiple initial obstacles, and determine the initial obstacles with a position change distance less than the first distance threshold as static obstacles.
[0066] For example, according to the position information of a certain initial obstacle in each image frame, the position change distance of the initial obstacle between adjacent image frames can be obtained, and the initial obstacles with a position change distance less than the first distance threshold are marked as static obstacles. For example, a vehicle parked by the roadside or a vehicle stopped due to waiting for a red light.
[0067] Because there may be errors in the tracking results of the 3D obstacle detection model, actually stationary obstacles may also show small movements in the tracking results. Therefore, the first distance threshold cannot be too small and is usually set to 0.1 meter.
[0068] Case 2. Determine the initial obstacles whose position information is detected only in one image frame among the multiple initial obstacles as static obstacles.
[0069] For example, an obstacle with a current track_id that appears only in one image frame is regarded as a static obstacle. Because there may be misdetections in the tracking results of the 3D obstacle detection model, resulting in additional obstacles, the misdetected obstacles need to be merged with the correctly detected static obstacles (in subsequent steps).
[0070] For example, a certain vehicle is detected in the first 50 frames and has a tracking number of track_101. In the 51st frame, due to misdetection, the vehicle is detected as another vehicle and assigned a tracking number of track_102, and then is detected as vehicle track_101 again. Then track_102 belongs to the obstacle that appears only in one image frame and needs to be regarded as a static obstacle and merged in subsequent steps to avoid the situation of tracking loss.
[0071] In some embodiments, determining the position information of the multiple static obstacles specifically includes:
[0072] For each static obstacle, the average value of the position information of at least one initial obstacle corresponding to the static obstacle is determined as the position information of the static obstacle. Since there may be errors in the tracking results and the positions of the same static obstacle in different image frames may be different, the position information of the initial obstacles corresponding to the same static obstacle in each graphic frame is averaged as the position information of the static obstacle.
[0073] In some embodiments, before step S130, it further includes:
[0074] Sort the position information of multiple static obstacles in descending order according to the number of corresponding position information, that is, sort by the number of times each static obstacle tracking number appears in how many image frames.
[0075] In some embodiments, as Figure 3 shown, step S130 specifically includes:
[0076] For each static obstacle, the static obstacles whose distance from the static obstacle is less than the second distance threshold are determined as the obstacles to be merged, and the position information of the static obstacle and the obstacles to be merged is merged to obtain the position information of the target obstacle, and then the position information of multiple target obstacles is obtained. Specifically, it includes the following steps:
[0077] S1301. Put multiple static obstacles into the first set.
[0078] For example, first put the sorted static obstacles into the first set as the set of obstacles to be processed.
[0079] S1302. Determine each static obstacle in the first set as the obstacle to be processed, and move the obstacle to be processed from the first set to the second set.
[0080] After determining each static obstacle in the first set as the obstacle to be processed, take out the i-th (initial value is 1) static obstacle and move it to the second set, and the second set is used as the merging set.
[0081] For example, move the first static obstacle track_001 in the first set to the second set as the obstacle to be processed.
[0082] S1303. Determine the static obstacles whose distance from the obstacle to be processed in the first set is less than the second distance threshold as the obstacles to be merged, and move the obstacles to be merged from the first set to the second set.
[0083] Calculate the distance between each remaining static obstacle in the first set and the obstacle to be processed in the second set. Determine the static obstacles with a distance less than the second distance threshold as the obstacles to be merged, and move the obstacles to be merged from the first set to the second set. Considering that there may be errors in the tracking results, the second distance threshold should not be too small and is usually set to 0.1 meters.
[0084] For example, starting from track_002 in the first set, calculate the distance between track_002 and track_001. If the distance is less than the second distance threshold, move track_002 to the second set; otherwise, it remains in the first set.
[0085] Then perform the same operation on track_003 until all static obstacles in the first set are traversed. For example, track_002, track_003, track_006, and track_007 are placed in the second set.
[0086] S1304. Merge the position information of the obstacles to be processed and the obstacles to be merged in the second set to obtain the position information of a target obstacle.
[0087] For example, merge the position information of the obstacle to be processed track_001 and the obstacles to be merged track_002, track_003, track_006, and track_007 in the second set, and consider them as the same static obstacle as track_001, thereby obtaining the position information of a target obstacle.
[0088] If different parts of a vehicle are detected as multiple obstacles, the merging of static obstacles can be achieved through the above steps. For example, if the body of a vehicle and multiple wheels are respectively detected as multiple obstacles, the body is track_001, and the multiple wheels are respectively track_002, track_003, track_006, and track_007. Through this step, they can be merged into a target obstacle to achieve the tracking and repair of static obstacles.
[0089] Another example is that since track_101 and track_102 are originally the same obstacle and their position information is the same or very close, they can also be merged into a target obstacle through the above steps to avoid the situation of tracking loss.
[0090] After that, initialize the second set, that is, clear the second set, and repeat steps S1302 to S1304. Move track_004 to the second set as the obstacle to be processed, and then start from track_005 in the first set to calculate the distance between track 005 and track_004. If the distance is less than the second distance threshold, move track005 to the second set; otherwise, it remains in the first set. Then, merge the position information of the obstacle to be processed track_004 and the obstacle to be merged in the second set, and take track_004 as the same static obstacle, so as to obtain the position information of a target obstacle.
[0091] And so on, repeat steps S1302 to S1304 until all the static obstacles in the first set are completely merged.
[0092] Corresponding to the method for processing vehicle obstacles provided by the embodiments of the present disclosure, the embodiments of the present disclosure also provide a device for processing vehicle obstacles. Figure 4 As shown in the structural block diagram of the device for processing vehicle obstacles provided by the embodiments of the present disclosure, Figure 4 the device for processing vehicle obstacles includes:
[0093] An acquisition module 41, configured to acquire a plurality of image frames collected for vehicle obstacles;
[0094] A determination module 42, configured to perform obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, where each of the image frames includes at least one of the static obstacles;
[0095] A merging module 43, configured to perform merging processing on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles.
[0096] In some embodiments, the determination module 42 is specifically configured to:
[0097] Perform obstacle detection on the plurality of image frames based on an obstacle detection model to determine the position information of a plurality of initial obstacles;
[0098] Extract a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles, and determine the position information of the plurality of static obstacles.
[0099] In some embodiments, extracting a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles includes at least one of the following:
[0100] Determine the initial obstacles whose position information is detected in only one image frame among the multiple initial obstacles as static obstacles;
[0101] Based on the position information of the multiple initial obstacles, determine the position change distance of each initial obstacle between adjacent image frames, and determine the initial obstacles with a position change distance less than the first distance threshold as static obstacles.
[0102] In some embodiments, the determining module 42 is further specifically configured to:
[0103] For each of the static obstacles, determine the average value of the position information of at least one initial obstacle corresponding to the static obstacle as the position information of the static obstacle.
[0104] In some embodiments, the merging module 43 is specifically configured to:
[0105] For each of the static obstacles, determine the static obstacles with a distance less than the second distance threshold from the static obstacle as the obstacles to be merged, and perform a merging process on the position information of the static obstacle and the obstacles to be merged to obtain the position information of a target obstacle, and further obtain the position information of multiple target obstacles.
[0106] In some embodiments, the merging module 43 is further specifically configured to:
[0107] Put the multiple static obstacles into a first set;
[0108] Determine each of the static obstacles in the first set as an obstacle to be processed, and move the obstacle to be processed from the first set to the second set;
[0109] Determine the static obstacles with a distance less than the second distance threshold from the obstacle to be processed in the first set as the obstacles to be merged, and move the obstacles to be merged from the first set to the second set;
[0110] Perform a merging process on the position information of the obstacle to be processed and the obstacles to be merged in the second set to obtain the position information of a target obstacle.
[0111] In some embodiments, the merging module 43 is further specifically configured to:
[0112] Before performing a merging process on the position information of the multiple static obstacles to obtain the position information of multiple target obstacles, sort the position information of the multiple static obstacles in descending order according to the corresponding number of position information.
[0113] Embodiments of the present disclosure also provide a computer-readable storage medium storing a program or instructions, which cause a computer to execute the steps of any of the above methods.
[0114] Calculate the position of the detected obstacle in the world coordinate system;
[0115] Filter out static obstacles from the detected obstacles;
[0116] Merge the static obstacles based on the spacing between the static obstacles;
[0117] Update the tracking data of the merged static obstacles.
[0118] Optionally, when executed by a computer processor, the computer-executable instructions can also be used to execute the technical solutions of any of the above methods for processing vehicle obstacles provided by the embodiments of the present disclosure, achieving corresponding beneficial effects.
[0119] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments of the present disclosure can be implemented by means of software and necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation. Based on this understanding, the technical solutions of the embodiments of the present disclosure, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disc of a computer, including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present disclosure.
[0120] Embodiments of the present disclosure also provide an electronic device, including: one or more processors; a memory for storing one or more programs or instructions; the processor is configured to execute the steps of any of the above methods by calling the programs or instructions stored in the memory, achieving corresponding beneficial effects.
[0121] Figure 5 It is a schematic hardware structure diagram of the electronic device provided by the embodiments of the present disclosure. As Figure 5 shown, the electronic device includes one or more processors 501 and a memory 502.
[0122] The processor 501 can be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and can control other components in the electronic device to execute desired functions.
[0123] The memory 502 may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 501 may run the program instructions to implement the method for processing vehicle obstacles in the embodiments of the present disclosure described above, and / or other desired functions. Various contents such as input signals, signal components, noise components, etc. may also be stored in the computer-readable storage medium.
[0124] In one example, the electronic device may further include: an input device 503 and an output device 504, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown).
[0125] In addition, the input device 503 may further include, for example, a keyboard, a mouse, etc.
[0126] The output device 504 may output various information to the outside, including the determined distance information, direction information, etc. The output device 504 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.
[0127] Of course, for simplicity, Figure 3 only some of the components related to the present disclosure in the electronic device are shown, and components such as buses, input / output interfaces, etc. are omitted. In addition, according to specific application scenarios, the electronic device may further include any other appropriate components.
[0128] It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device including the said element.
[0129] The above are only specific embodiments of the present disclosure, enabling those skilled in the art to understand or implement the present disclosure. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to these embodiments described herein, but rather will be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for processing vehicle obstacles, characterized in that, Including: Obtaining a plurality of image frames collected for vehicle obstacles; Performing obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, wherein each of the image frames includes at least one of the static obstacles; Performing a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles.
2. The method according to claim 1, characterized in that, Performing obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, including: Performing obstacle detection on the plurality of image frames based on an obstacle detection model to determine the position information of a plurality of initial obstacles; Extracting a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles and determining the position information of the plurality of static obstacles.
3. The method according to claim 2, characterized in that Extracting a plurality of static obstacles from the plurality of initial obstacles based on the position information of the plurality of initial obstacles, including at least one of the following: Determining an initial obstacle whose position information is detected only in one image frame among the plurality of initial obstacles as a static obstacle; Determining the position change distance of each of the initial obstacles between adjacent image frames based on the position information of the plurality of initial obstacles and determining an initial obstacle with a position change distance less than a first distance threshold as a static obstacle.
4. The method according to claim 2, wherein The determining the position information of the plurality of static obstacles includes: For each of the static obstacles, determining the average value of the position information of at least one initial obstacle corresponding to the static obstacle as the position information of the static obstacle.
5. The method according to claim 1, characterized in that, Performing a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles, including: For each of the static obstacles, determining a static obstacle whose distance from the static obstacle is less than a second distance threshold as an obstacle to be merged, and performing a merging process on the position information of the static obstacle and the obstacle to be merged to obtain the position information of a target obstacle, and further obtaining the position information of a plurality of target obstacles.
6. The method according to claim 5, characterized in that For each of the static obstacles, determining a static obstacle whose distance from the static obstacle is less than a second distance threshold as an obstacle to be merged, and performing a merging process on the position information of the static obstacle and the obstacle to be merged to obtain the position information of a corresponding target obstacle, including: Putting the plurality of static obstacles into a first set; Determining each of the static obstacles in the first set as an obstacle to be processed and moving the obstacle to be processed from the first set to a second set; Determining a static obstacle whose distance from the obstacle to be processed in the first set is less than the second distance threshold as an obstacle to be merged and moving the obstacle to be merged from the first set to the second set; Performing a merging process on the position information of the obstacle to be processed and the obstacle to be merged in the second set to obtain the position information of a target obstacle.
7. The method according to claim 1, characterized in that Before performing a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles, the method further includes: Sort the position information of the multiple static obstacles in descending order according to the number of corresponding position information.
8. A processing device for vehicle obstacles, characterized in that, Comprising: An acquisition module, configured to acquire a plurality of image frames collected for vehicle obstacles; A determination module, configured to perform obstacle detection on the plurality of image frames to determine the position information of a plurality of static obstacles, wherein each of the image frames includes at least one of the static obstacles; A merging module, configured to perform a merging process on the position information of the plurality of static obstacles to obtain the position information of a plurality of target obstacles.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program or instruction, and the program or instruction causes a computer to execute the steps of the method according to any one of claims 1 to 7.
10. An electronic device, characterized in that, Comprising: One or more processors; A memory, configured to store one or more programs or instructions; The processor is configured to execute the steps of the method according to any one of claims 1 to 7 by calling the program or instruction stored in the memory.