Rear crossing traffic early warning method and device, electronic equipment and storage medium

Abstract

The invention provides a rear crossing traffic early warning method and device, electronic equipment and a storage medium, and the method comprises the steps: collecting an original image of each camera on a vehicle, and obtaining multiple paths of original images; carrying out perspective transformation on the original image into an image at a top view angle; performing target detection and semantic segmentation on each path of image to obtain target information of traffic participants and semantic segmentation information of a drivable area; based on the calibration parameters of the camera, performing inverse perspective mapping on the target information of the traffic participants in each path of image and the semantic segmentation information of the drivable area to an aerial view coordinate system with the vehicle as the center, and generating the tracking state information of the traffic participants and a static map of the drivable area; calculating a collision risk parameter between the vehicle and the traffic participant based on the backing track of the vehicle, the tracking state information and the travelable area static map; and when the collision risk parameter satisfies an early warning condition, sending an early warning instruction. Through the method, the cost is reduced, and the target identification capability is improved.

Description

Technical Field

[0001] This application relates to the field of automotive technology, and in particular to a rear cross-traffic warning method, device, electronic device, and storage medium. Background Technology

[0002] Rear Cross Traffic Alert (RCTA), also known as rear cross traffic warning, is an important component of modern automotive active safety systems. It primarily monitors moving targets such as vehicles and pedestrians in the blind spots to the sides and rear when a vehicle is reversing, and issues a warning to the driver when a collision risk is present. Currently, the mainstream RCTA technology is based on millimeter-wave radar. However, this approach suffers from several drawbacks, including high cost (millimeter-wave radar hardware is expensive, increasing overall vehicle manufacturing costs and hindering its widespread adoption in economy vehicles), limited target recognition capabilities (radar performs poorly in recognizing static targets and struggles to distinguish target types such as vehicles, pedestrians, and obstacles), insufficient environmental adaptability (radar signals may be blocked when a vehicle is parked between large vehicles (such as trucks or buses), leading to detection failure), and installation limitations (radar requires specific installation locations and angles, placing high demands on the overall vehicle layout). Summary of the Invention

[0003] In view of this, the purpose of this application is to provide a rear cross-traffic warning method, device, electronic device and storage medium to reduce costs and installation difficulty, while improving target recognition capability and environmental adaptability.

[0004] In a first aspect, embodiments of this application provide a method for warning of rear-facing cross-traffic, including: Simultaneously acquire raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; Perform perspective transformation on each original image to transform each original image into a top-down view image; Target detection and semantic segmentation are performed on each image from the top-down view to obtain target information of traffic participants and semantic segmentation information of drivable areas in each image. Based on the calibration parameters of each camera, the target information of the traffic participants and the semantic segmentation information of the drivable area in each image are inversely perspective mapped to a unified bird's-eye view coordinate system centered on the vehicle. The same traffic participant mapped from different cameras to the bird's-eye view coordinate system is matched and fused through a data association algorithm. A global ID is assigned to each traffic participant and target tracking is performed to generate the tracking status information of the traffic participants. The semantic segmentation information of the mapped drivable area is also fused to generate a static map of the drivable area under the bird's-eye view coordinate system. The vehicle's reversing trajectory is predicted based on its own vehicle status information, and the collision risk parameters between the vehicle and each traffic participant are calculated based on the reversing trajectory, the tracking status information of traffic participants in the bird's-eye view coordinate system, and the static map of the drivable area. The collision risk parameters are used to determine whether the warning conditions are met. When the warning conditions are met, a warning command is generated and sent.

[0005] In conjunction with the first aspect, embodiments of this application provide a first possible implementation of the first aspect, wherein performing perspective transformation on each original image to transform each original image into a top-down view image includes: Lens distortion correction is performed on each original image to obtain the distortion-corrected image; A perspective transformation is performed on each distortion-corrected image to transform it into a top-down view.

[0006] In conjunction with the first aspect, this application provides a second possible implementation of the first aspect, wherein the collision risk parameters include: estimated collision time and minimum predicted distance; the step of determining whether a warning condition is met based on the collision risk parameters, and generating and sending a warning command when the warning condition is met, includes: When the predicted collision time is less than the first time threshold and the minimum predicted distance is less than the first distance threshold, it is determined that the first-level warning condition is met, and a first-level warning instruction is generated and sent. When the estimated collision time is less than the second time threshold and the minimum predicted distance is less than the second distance threshold, it is determined that the conditions for a level 2 warning are met, and a level 2 warning instruction is generated and sent; wherein, the first time threshold is greater than the second time threshold, the first distance threshold is greater than the second distance threshold, and the urgency of the level 1 warning is lower than that of the level 2 warning.

[0007] In conjunction with the second possible implementation of the first aspect, this application provides a third possible implementation of the first aspect, wherein the first time threshold is 3 seconds, the first distance threshold is 4 meters, the second time threshold is 1.5 seconds, and the second distance threshold is 2 meters.

[0008] In conjunction with the first aspect, this application provides a fourth possible implementation of the first aspect, wherein the camera further includes a 360-degree surround view camera; the horizontal field of view of the left rearview camera and the right rearview camera is not less than 60 degrees, and the lateral distance between the lower edge of their vertical field of view and the wheel of the vehicle is not greater than 2.5 meters.

[0009] In conjunction with the first aspect, this application provides a fifth possible implementation of the first aspect, wherein the vehicle status information includes: the vehicle's steering wheel angle, vehicle speed, and gear information.

[0010] In conjunction with the first aspect, this application provides a sixth possible implementation of the first aspect, wherein sending the warning instruction includes: Warning commands are sent to the instrument panel or audio head unit via the vehicle's CAN bus to trigger acoustic and / or optical alarms.

[0011] Secondly, embodiments of this application also provide a rear cross-traffic warning device, comprising: The acquisition module is used to simultaneously acquire raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; The transformation module is used to perform perspective transformation on each original image to transform each original image into a top-down view image; The detection module is used to perform target detection and semantic segmentation on each image from the top-down view, so as to obtain the target information of traffic participants and the semantic segmentation information of the drivable area in each image. The mapping module is used to inversely perspective map the target information of the traffic participants and the semantic segmentation information of the drivable area in each image to a unified bird's-eye view coordinate system centered on the vehicle, based on the calibration parameters of each camera. The module matches and fuses the same traffic participants mapped from different cameras to the bird's-eye view coordinate system through a data association algorithm, assigns a global ID to each traffic participant and performs target tracking, generates tracking status information of the traffic participants, and fuses the semantic segmentation information of the mapped drivable area to generate a static map of the drivable area in the bird's-eye view coordinate system. The calculation module is used to predict the reversing trajectory of the vehicle based on the vehicle's own vehicle status information, and to calculate the collision risk parameters between the vehicle and each traffic participant based on the reversing trajectory, the tracking status information of traffic participants in the bird's-eye view coordinate system and the static map of the drivable area. The early warning module is used to determine whether the early warning conditions are met based on the collision risk parameters, and to generate and send an early warning command when the early warning conditions are met.

[0012] Thirdly, embodiments of this application also provide an electronic device, including: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory via the bus, and when the machine-readable instructions are executed by the processor, the steps in any of the possible implementations of the first aspect described above are performed.

[0013] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps in any of the possible implementations of the first aspect described above.

[0014] This application provides a rear cross-traffic warning method, device, electronic device, and storage medium. It utilizes cameras (left-side rearview camera, right-side rearview camera, and rearview camera) instead of expensive millimeter-wave radar as the sensing hardware, significantly reducing costs. Through target detection and semantic segmentation, it can identify the types of traffic participants (such as vehicles and pedestrians) and drivable areas, overcoming the limitation of radar in distinguishing target categories. Furthermore, the optical imaging-based cameras are not affected by metal objects obstructing radar signals and can still operate in complex scenarios such as between vehicles, solving the problem of insufficient environmental adaptability. Moreover, compared to radar, camera placement is more flexible, typically reusing the installation points and layout of existing vision systems, reducing the complexity of vehicle-wide deployment.

[0015] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1A flowchart of a rear cross-traffic warning method provided in an embodiment of this application is shown; Figure 2 This illustration shows a structural schematic diagram of a rear cross-traffic warning device provided in an embodiment of this application; Figure 3 A schematic diagram of the structure of an electronic device provided in an embodiment of this application is shown. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0019] Rear Cross Traffic Alert (RCTA), a crucial component of modern automotive active safety systems, primarily monitors moving targets such as vehicles and pedestrians in the blind spots to the sides and rear when the vehicle is reversing, and issues a warning to the driver when a collision risk is present. Currently, the mainstream RCTA technology is based on millimeter-wave radar. However, this approach has the following significant drawbacks: 1. High cost: Millimeter-wave radar hardware is expensive, which increases the overall vehicle manufacturing cost and makes it difficult to popularize in economy vehicles.

[0020] 2. Limited target recognition capability: Radar has poor recognition performance for static targets and has difficulty distinguishing target types (such as vehicles, pedestrians, and obstacles).

[0021] 3. Insufficient environmental adaptability: When a vehicle is parked between large vehicles (such as trucks or buses), the radar signal may be blocked, causing detection failure.

[0022] 4. Installation limitations: Radar requires specific installation positions and angles, which places high demands on the overall vehicle layout.

[0023] Based on this, embodiments of this application provide a rear cross-traffic warning method, device, electronic device, and storage medium, which are described below through embodiments.

[0024] To facilitate understanding of this embodiment, a method for warning of rear-crossing traffic disclosed in this application will first be described in detail. For example... Figure 1 As shown, the process includes the following steps S101-S106: S101: Synchronously acquires raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; S102: Perform perspective transformation on each original image to transform each original image into a top-down view image; S103: Perform target detection and semantic segmentation on each image from the top-down view to obtain target information of traffic participants and semantic segmentation information of drivable areas in each image. S104: Based on the calibration parameters of each camera, the target information of traffic participants and the semantic segmentation information of drivable areas in each image are inversely perspective mapped to a unified bird's-eye view coordinate system centered on the vehicle. The same traffic participant mapped to the bird's-eye view coordinate system by different cameras is matched and fused through a data association algorithm. A global ID is assigned to each traffic participant and target tracking is performed to generate the tracking status information of the traffic participants. The semantic segmentation information of the mapped drivable areas is fused to generate a static map of the drivable areas under the bird's-eye view coordinate system. S105: Predict the vehicle's reversing trajectory based on the vehicle's own vehicle status information, and calculate the collision risk parameters between the vehicle and each traffic participant based on the reversing trajectory, the tracking status information of traffic participants in the bird's-eye view coordinate system, and the static map of the drivable area. S106: Determine whether the warning conditions are met based on the collision risk parameters. If the warning conditions are met, generate and send a warning command.

[0025] In step S101, the side and rearview cameras installed on the vehicle (including the left and right rearview cameras) have the following characteristics: 3 megapixels, resolution ≥1920*1536, HFOV 60°, Yaw angle 150° (reference is the ideal plane of the vehicle body), no interference above horizontal, VFOV ≥15°, and the lateral distance between the lower edge of the VFOV and the vehicle wheel ≤2.5m.

[0026] Rearview camera: 3 megapixels, resolution ≥1920*1536, HFOV 60°, horizontal and above interference-free VFOV ≥15°, distance between the lower edge of the VFOV and the rear of the vehicle ≤6m, centered.

[0027] The camera system also includes 360-degree surround view cameras, which are arranged according to the requirements of panoramic vehicle imaging.

[0028] In step S102, lens distortion correction is performed on each original image to obtain a distortion-corrected image; perspective transformation is performed on each distortion-corrected image to transform it into a top-down view image.

[0029] In step S103, the multi-channel images from the top-down perspective are input into the target detection and segmentation model. The target detection and segmentation model performs target detection and semantic segmentation on each channel image from the top-down perspective to obtain the target information of traffic participants and the semantic segmentation information of the drivable area in each channel image.

[0030] In step S104, based on the calibration parameters of each camera, the target information of traffic participants and the semantic segmentation information of drivable areas in each image are inversely perspective mapped to a unified bird's-eye view coordinate system centered on the vehicle. The same traffic participant mapped to the bird's-eye view coordinate system by different cameras is matched and fused using a data association algorithm (Hungarian algorithm). The SORT / DeepSORT algorithm is used to assign a global ID to each traffic participant and perform target tracking to generate the tracking status information of the traffic participants. The semantic segmentation information of the mapped drivable areas is also fused to generate a static map of the drivable areas in the bird's-eye view coordinate system. In step S105, the vehicle status information includes: the vehicle's steering wheel angle, speed, and gear information. The static map of the drivable area is used to exclude traffic participants located outside the road area when calculating collision risk parameters.

[0031] In step S106, the collision risk parameters include: expected collision time and minimum predicted distance.

[0032] When the expected collision time is less than the first time threshold and the minimum predicted distance is less than the first distance threshold, the conditions for a Level 1 warning are met, and a Level 1 warning command is generated and sent. When the expected collision time is less than the second time threshold and the minimum predicted distance is less than the second distance threshold, the conditions for a level 2 warning are met, and a level 2 warning instruction is generated and sent. Among these conditions, the first time threshold is greater than the second time threshold, the first distance threshold is greater than the second distance threshold, and the urgency level of a level 1 warning is lower than that of a level 2 warning.

[0033] In one specific embodiment, the first time threshold is 3 seconds and the first distance threshold is 4 meters; the second time threshold is 1.5 seconds and the second distance threshold is 2 meters.

[0034] When sending the warning command in step S106, the specific steps can be followed: Warning commands are sent to the instrument panel or audio head unit via the vehicle's CAN bus to trigger acoustic and / or optical alarms.

[0035] Based on the same technical concept, embodiments of this application also provide a rear cross-traffic warning device, such as... Figure 2 As shown, the device includes: The acquisition module 201 is used to simultaneously acquire the raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; Transformation module 202 is used to perform perspective transformation on each original image to transform each original image into an image from a top-down perspective; The detection module 203 is used to perform target detection and semantic segmentation on each image from the top-down view, so as to obtain the target information of traffic participants and the semantic segmentation information of the drivable area in each image. The mapping module 204 is used to inversely perspective map the target information of the traffic participants and the semantic segmentation information of the drivable area in each image to a unified bird's-eye view coordinate system centered on the vehicle, based on the calibration parameters of each camera. The module then matches and fuses the same traffic participants mapped from different cameras to the bird's-eye view coordinate system using a data association algorithm. It assigns a global ID to each traffic participant and performs target tracking, generates tracking status information of the traffic participants, and fuses the semantic segmentation information of the mapped drivable area to generate a static map of the drivable area in the bird's-eye view coordinate system. The calculation module 205 is used to predict the reversing trajectory of the vehicle based on the vehicle's own status information, and to calculate the collision risk parameters between the vehicle and each traffic participant based on the reversing trajectory, the tracking status information of traffic participants in the bird's-eye view coordinate system and the static map of the drivable area. The early warning module 206 is used to determine whether the early warning conditions are met based on the collision risk parameters, and to generate and send an early warning command when the early warning conditions are met.

[0036] Optionally, when the transformation module 202 performs perspective transformation on each original image to transform each original image into a top-down view image, it is specifically used for: Lens distortion correction is performed on each original image to obtain the distortion-corrected image; A perspective transformation is performed on each distortion-corrected image to transform it into a top-down view.

[0037] Optionally, the collision risk parameters include: estimated collision time and minimum predicted distance; when the early warning module 306 is used to determine whether the early warning conditions are met based on the collision risk parameters, and to generate and send an early warning command when the early warning conditions are met, it is specifically used for: When the predicted collision time is less than the first time threshold and the minimum predicted distance is less than the first distance threshold, it is determined that the first-level warning condition is met, and a first-level warning instruction is generated and sent. When the estimated collision time is less than the second time threshold and the minimum predicted distance is less than the second distance threshold, it is determined that the conditions for a level 2 warning are met, and a level 2 warning instruction is generated and sent; wherein, the first time threshold is greater than the second time threshold, the first distance threshold is greater than the second distance threshold, and the urgency of the level 1 warning is lower than that of the level 2 warning.

[0038] Optionally, the first time threshold is 3 seconds and the first distance threshold is 4 meters; the second time threshold is 1.5 seconds and the second distance threshold is 2 meters.

[0039] Optionally, the camera also includes a 360-degree surround view camera; the horizontal field of view of the left rearview camera and the right rearview camera is not less than 60 degrees, and the lateral distance between the lower edge of their vertical field of view and the wheel of the vehicle is not greater than 2.5 meters.

[0040] Optionally, the vehicle status information includes: the vehicle's steering wheel angle, vehicle speed, and gear information.

[0041] Optionally, when the early warning module 206 is used to send early warning commands, it is specifically used for: Warning commands are sent to the instrument panel or audio head unit via the vehicle's CAN bus to trigger acoustic and / or optical alarms.

[0042] Figure 3 A schematic diagram of an electronic device provided in this application embodiment includes: a processor 301, a memory 302, and a bus 303. The memory 302 stores machine-readable instructions executable by the processor 301. When the electronic device runs the above-described information processing method, the processor 301 and the memory 302 communicate through the bus 303. The processor 301 executes the machine-readable instructions to perform the steps of the method described in Embodiment 1.

[0043] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps described in Embodiment 1.

[0044] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, electronic devices, and computer-readable storage media described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0045] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, electronic devices, and computer-readable storage media can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or modules may be electrical, mechanical, or other forms.

[0046] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0047] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0048] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0049] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The scope of protection of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of protection of the claims.

Claims

1. A method for early warning of traffic crossing from behind, characterized in that, include: Simultaneously acquire raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; Perform perspective transformation on each original image to transform each original image into a top-down view image; Target detection and semantic segmentation are performed on each image from the top-down view to obtain target information of traffic participants and semantic segmentation information of drivable areas in each image. Based on the calibration parameters of each camera, the target information of the traffic participants and the semantic segmentation information of the drivable area in each image are inversely perspective mapped to a unified bird's-eye view coordinate system centered on the vehicle. The same traffic participant mapped from different cameras to the bird's-eye view coordinate system is matched and fused through a data association algorithm. A global ID is assigned to each traffic participant and target tracking is performed to generate the tracking status information of the traffic participants. The semantic segmentation information of the mapped drivable area is also fused to generate a static map of the drivable area under the bird's-eye view coordinate system. The vehicle's reversing trajectory is predicted based on its own vehicle status information, and the collision risk parameters between the vehicle and each traffic participant are calculated based on the reversing trajectory, the tracking status information of traffic participants in the bird's-eye view coordinate system, and the static map of the drivable area. The collision risk parameters are used to determine whether the warning conditions are met. When the warning conditions are met, a warning command is generated and sent.

2. The method according to claim 1, characterized in that, The step of performing perspective transformation on each original image to transform each original image into a top-down view includes: Lens distortion correction is performed on each original image to obtain the distortion-corrected image; A perspective transformation is performed on each distortion-corrected image to transform it into a top-down view.

3. The method according to claim 1, characterized in that, The collision risk parameters include: estimated collision time and minimum predicted distance; the step of determining whether the warning conditions are met based on the collision risk parameters, and generating and sending a warning command when the warning conditions are met, includes: When the predicted collision time is less than the first time threshold and the minimum predicted distance is less than the first distance threshold, it is determined that the first-level warning condition is met, and a first-level warning instruction is generated and sent. When the estimated collision time is less than the second time threshold and the minimum predicted distance is less than the second distance threshold, it is determined that the conditions for a level 2 warning are met, and a level 2 warning instruction is generated and sent; wherein, the first time threshold is greater than the second time threshold, the first distance threshold is greater than the second distance threshold, and the urgency of the level 1 warning is lower than that of the level 2 warning.

4. The method according to claim 3, characterized in that, The first time threshold is 3 seconds, and the first distance threshold is 4 meters; the second time threshold is 1.5 seconds, and the second distance threshold is 2 meters.

5. The method according to claim 1, characterized in that, The camera also includes a 360-degree surround view camera; the horizontal field of view of the left rearview camera and the right rearview camera is not less than 60 degrees, and the lateral distance between the lower edge of their vertical field of view and the wheel of the vehicle is not greater than 2.5 meters.

6. The method according to claim 1, characterized in that, The vehicle status information includes: the vehicle's steering wheel angle, speed, and gear information.

7. The method according to claim 1, characterized in that, The sending of the warning instruction includes: Warning commands are sent to the instrument panel or audio head unit via the vehicle's CAN bus to trigger acoustic and / or optical alarms.

8. A rear-crossing traffic warning device, characterized in that, include: The acquisition module is used to simultaneously acquire raw images captured by each camera on the vehicle to obtain multiple raw images; the cameras include a left rearview camera, a right rearview camera, and a rearview camera; The transformation module is used to perform perspective transformation on each original image to transform each original image into a top-down view image; The detection module is used to perform target detection and semantic segmentation on each image from the top-down view, so as to obtain the target information of traffic participants and the semantic segmentation information of the drivable area in each image. The mapping module is used to inversely perspective map the target information of the traffic participants and the semantic segmentation information of the drivable area in each image to a unified bird's-eye view coordinate system centered on the vehicle, based on the calibration parameters of each camera. The module matches and fuses the same traffic participants mapped from different cameras to the bird's-eye view coordinate system through a data association algorithm, assigns a global ID to each traffic participant and performs target tracking, generates tracking status information of the traffic participants, and fuses the semantic segmentation information of the mapped drivable area to generate a static map of the drivable area in the bird's-eye view coordinate system. The calculation module is used to predict the reversing trajectory of the vehicle based on the vehicle's own vehicle status information, and to calculate the collision risk parameters between the vehicle and each traffic participant based on the tracking status information of traffic participants in the bird's-eye view coordinate system and the static map of the drivable area. The early warning module is used to determine whether the early warning conditions are met based on the collision risk parameters, and to generate and send an early warning command when the early warning conditions are met.

9. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the memory via the bus, and the machine-readable instructions, when executed by the processor, perform the steps of the method as described in 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 that, when executed by a processor, performs the steps of the method as described in any one of claims 1 to 7.

Images