Vehicle collision avoidance warning methods and vehicles

Abstract

This application provides a vehicle collision avoidance warning method and a vehicle, relating to the field of intelligent sensing technology. The method includes: when a visual obstruction area is detected in front of the vehicle, acquiring reflected images of other target vehicles based on the visual obstruction area, wherein the reflected images represent visual images reflected from the visual obstruction area onto other target vehicles; responding to the presence of a target object in the reflected images, predicting the current collision risk between the vehicle and the target object based on the current motion information of the target object, wherein the target object represents an object in the visual obstruction area that obstructs the vehicle's movement; and controlling the vehicle to issue a collision avoidance warning based on the current collision risk. Based on the above scheme, a collision can be avoided when the area in front of the vehicle is obstructed.

Description

Technical Field

[0001] This application relates to the field of intelligent sensing technology, and more particularly to a vehicle collision avoidance warning method and vehicle in the field of intelligent sensing technology. Background Technology

[0002] When a vehicle is in motion, it can capture images of the area in front of it and use these images to detect obstacles. Upon detecting an obstacle, the vehicle can issue a collision avoidance warning in a timely manner, effectively preventing a collision. However, if the area in front of the vehicle is obstructed, the vehicle may be unable to detect obstacles in the obstructed area, thus failing to trigger a collision avoidance warning in time. This increases the likelihood of a collision between the vehicle and the obstacle, reducing driving safety.

[0003] Therefore, how to avoid vehicle collisions when the front of the vehicle is obstructed is an urgent problem that needs to be solved. Summary of the Invention

[0004] This application provides a vehicle collision avoidance warning method and a vehicle, which can prevent a vehicle from colliding when the front of the vehicle is obstructed.

[0005] Firstly, this application provides a vehicle collision avoidance warning method, the method comprising: When a visual obstruction area is detected in front of the vehicle, the reflected images on other target vehicles are acquired based on the visual obstruction area. The reflected images are used to represent the visual images reflected from the visual obstruction area to other target vehicles. The visual obstruction area is caused by external obstructions of the vehicle. In response to the presence of a target object in the reflected image, the current collision risk between the vehicle and the target object is predicted based on the current motion information of the target object, wherein the target object is used to represent an object that obstructs the vehicle's movement in the visual occlusion area; Based on the current collision risk, control the vehicle to issue a collision avoidance warning.

[0006] In this embodiment, when a visual obstruction area is detected in front of the vehicle, in order to predict the collision risk between the vehicle and an object (i.e., a target object) obstructing the vehicle's movement within the visual obstruction area, reflected images from other target vehicles can be determined through the visual obstruction area. The current motion information of the target object in the reflected image is then used to predict the current collision risk between the vehicle and the target object, thereby controlling the vehicle to issue a collision avoidance warning. Compared to the problem of being unable to detect target objects within a visual obstruction area in front of the vehicle, thus preventing collision risk prediction, this embodiment utilizes the principle of optical reflection to obtain reflected images from other target vehicles through the visual obstruction area. This indirectly detects target objects that are not within the vehicle's visual range. Upon detection of the target object, the current motion information of the target object is used to predict the collision risk between the vehicle and the target object. This predicted collision risk allows for early collision avoidance warnings, minimizing the possibility of collisions and effectively improving vehicle safety.

[0007] In conjunction with the first aspect, in certain implementations of the first aspect, the acquisition of reflected images on other target vehicles based on visual occlusion areas includes: Get the current distances between other target vehicles and between vehicles; Based on the visual occlusion area and the current distance, the target reflection area on other target vehicles is determined, wherein the area size of the target reflection area is positively correlated with the area size of the visual occlusion area and the area size of the target reflection area is positively correlated with the current distance; The reflected image is determined based on visual information in the target's reflective area.

[0008] In this embodiment, when determining the target reflection area on other target vehicles, the determination is made by combining the visual occlusion area of ​​the vehicle with the current distance between the other target vehicles. This takes into account the influence of the current distance between the other target vehicles on the target reflection area, avoiding potential biases that may occur when determining the target reflection area solely based on the visual occlusion area of ​​the vehicle, thus improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for the vehicle and the target object, thereby further preventing collisions between the vehicle and the target object.

[0009] In conjunction with the first aspect and the above implementation methods, in some implementation methods of the first aspect, the determination of the target reflection area on other target vehicles based on the visual occlusion area and the current distance includes: Based on the visual occlusion area and the current distance, the initial reflection area on other target vehicles is determined, wherein the area size of the visual occlusion area is positively correlated with the area size of the initial reflection area, and the current distance is positively correlated with the area size of the initial reflection area; The initial reflection area is adjusted based on the current road slope of other target vehicles to obtain the target reflection area.

[0010] In this embodiment, when determining the target reflection area on other target vehicles, the current road slope of the other target vehicles is also considered in addition to the vehicle's visual occlusion area and the current distance between other target vehicles. This avoids potential deviations when determining the target reflection area based solely on the vehicle's visual occlusion area and the current distance between other target vehicles, further improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for vehicles and target objects, thereby further preventing collisions between vehicles and target objects.

[0011] Combining the first aspect and the above-described implementation methods, in some implementation methods of the first aspect, the adjustment of the initial reflection area size based on the current road slope of other target vehicles to obtain the target reflection area includes: Based on the current road slope, determine the target adjustment amount, where the target adjustment amount is positively correlated with the current road slope; Based on the target adjustment amount, the initial reflection area is adjusted to obtain the target reflection area.

[0012] In this embodiment, the reflection areas determined by the visual occlusion area of ​​passing vehicles, the current distance between other target vehicles, and the current road slope of other target vehicles are adjusted to obtain the target reflection areas on other target vehicles, further improving the accuracy of the target reflection areas. Furthermore, with more accurate target reflection areas, more accurate collision risk prediction can be made for vehicles and target objects, thereby further avoiding collisions between vehicles and target objects.

[0013] Combining the first aspect and the above implementation methods, in some implementation methods of the first aspect, the initial reflection region is adjusted based on the target adjustment amount to obtain the target reflection region, including: Based on the target adjustment amount, the initial reflection area is adjusted to obtain the adjusted reflection area; The region of interest in the adjusted reflection region is determined as the target reflection region; Among them, the probability of visual information in the visually occluded area of ​​the region of interest is greater than the probability of visual information in the visually occluded area of ​​the non-region of interest in the adjusted reflection area.

[0014] In this embodiment, the reflection area determined by the adjustment amount based on the current road slope of other target vehicles is adjusted to the visual occlusion area of ​​passing vehicles, the current distance between other target vehicles, and the reflection area determined by the current distance between vehicles, thereby obtaining an adjusted reflection area and improving the accuracy of the adjusted reflection area. Furthermore, based on the more accurate adjusted reflection area, a more accurate target reflection area can be obtained.

[0015] Furthermore, defining the region of interest within the adjusted reflection region as the target reflection region avoids interference from non-interest regions on the target reflection region results, further improving the accuracy of the target reflection region.

[0016] Furthermore, by selecting the region of interest, the amount of data for subsequent processing can be effectively reduced. It is not necessary to perform full calculations on the adjusted reflection region; only partial calculations on the region of interest are required. This allows collision risk prediction to focus on the region of interest, thereby reducing the computational load of collision risk prediction.

[0017] In conjunction with the first aspect and the above-described implementations, in some implementations of the first aspect, the acquisition of reflected images on other target vehicles based on visual occlusion areas includes: Based on the geometric dimensions of the visual occlusion area, the target reflection area on other target vehicles is determined. The geometric dimensions of the target reflection area are positively correlated with the geometric dimensions of the visual occlusion area, and there is a corresponding relationship between the geometric dimensions of the visual occlusion area and the target reflection area. The reflected image is determined based on visual information in the target's reflective area.

[0018] In this embodiment, when determining the target reflection area on other target vehicles, the geometric dimensions of the visual occlusion area are used for determination. This ensures a stable mapping relationship between the determined target reflection area and the geometric dimensions of the visual occlusion area, thereby improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for vehicles and target objects, thereby further avoiding collisions between vehicles and target objects.

[0019] In conjunction with the first aspect and the above-described implementations, in some implementations of the first aspect, the prediction of the current collision risk between the vehicle and the target object based on the target object's current motion information includes: Based on the target object's current motion information, determine the target object's current motion intention; Predict the current collision risk based on the target object's current movement intention.

[0020] In this embodiment of the application, the current motion intention of the target object is determined by the current motion information of the target object, and the future motion trend of the target object can be predicted. In order to predict the current collision risk between the vehicle and the target object in advance based on the current motion intention of the target object, sufficient warning time is reserved for the vehicle, thereby improving the timeliness and reliability of the collision warning.

[0021] Combining the first aspect and the above implementation methods, in some implementation methods of the first aspect, the prediction of the current collision risk based on the current movement intention of the target object includes: Based on the vehicle's current operating information, determine the vehicle's current driving intention; Based on the target object's current motion intention and the vehicle's current driving intention, predict the current collision risk.

[0022] In conjunction with the first aspect and the above-described implementations, in some implementations of the first aspect, the prediction of the current collision risk between the vehicle and the target object based on the target object's current motion information includes: Based on the target object's current motion information, determine the target object's current motion intention; Based on the vehicle's current operating information, determine the vehicle's current driving intention; Based on the target object's current motion intention and the vehicle's current driving intention, predict the current collision risk.

[0023] In this embodiment, the collision risk prediction between the vehicle and the target object is jointly performed based on the target object's current motion intention and the vehicle's current driving intention. This takes into account the impact of the vehicle's current driving intention on the collision risk prediction result, avoiding potential biases that might occur when predicting collision risk solely based on the target object's current motion intention, and improving the accuracy of the collision risk prediction result. Furthermore, with a more accurate collision risk prediction result, collisions between the vehicle and the target object are further avoided.

[0024] Furthermore, by determining the vehicle's current driving intention through its current operating information, the future movement trend of the vehicle can be predicted. This allows for the prediction of the current collision risk between the vehicle and the target object based on the target object's current movement intention and the vehicle's current driving intention, thus providing sufficient warning time for the vehicle and further improving the timeliness and reliability of collision warnings.

[0025] Combining the first aspect and the above-described implementation methods, in some implementation methods of the first aspect, the prediction of the current collision risk based on the current motion intention of the target object and the current driving intention of the vehicle includes: When there is only one target object, the current collision risk is predicted based on the current motion intention of the target object and the current driving intention of the vehicle. When there are multiple target objects, a first target object is determined based on the current distance between each target object and the vehicle; the current collision risk is predicted based on the current movement intention of the first target object and the current driving intention of the vehicle, where the current distance between the first target object and the vehicle is the smallest; or, When there are multiple target objects, the collision risk between the vehicle and each target object is determined based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle. The collision risks of multiple targets are fused to obtain the current collision risk.

[0026] In this embodiment, selecting different collision risk prediction methods based on the number of target objects makes collision risk prediction between vehicles and target objects more flexible and more consistent with the actual situation of the target objects, further improving the accuracy of collision risk prediction results. Consequently, based on more accurate collision risk prediction results, collisions between vehicles and target objects can be further avoided.

[0027] Combining the first aspect and the above implementation methods, in some implementation methods of the first aspect, the determination of the target collision risk between the vehicle and each target object based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle includes: Based on the current motion intention of each target object and the current driving intention of the vehicle, the initial collision risk between the vehicle and each target object is determined; Based on the current distance between each target object and the vehicle, the weight of each target object is determined, wherein the weight of each target object is negatively correlated with the current distance between each target object and the vehicle; The initial collision risk between the vehicle and each target object is weighted by the weight of each target object to obtain the target collision risk between the vehicle and each target object.

[0028] In this embodiment, when determining the collision risk between a vehicle and a target object, the impact of the current distance between the target object and the vehicle on the collision risk is considered. The collision risk is adjusted using a weight determined by the current distance between the target object and the vehicle, which improves the accuracy of the collision risk prediction results. Furthermore, based on more accurate collision risk prediction results, collisions between the vehicle and the target object can be further avoided.

[0029] Secondly, this application provides a vehicle collision avoidance warning device, the device comprising: The acquisition module is used to acquire reflected images on other target vehicles based on the visual occlusion area when a visual occlusion area is detected in front of the vehicle. The reflected images are used to represent the visual images reflected from the visual occlusion area to other target vehicles. The visual occlusion area is caused by external obstructions of the vehicle. The processing module is used to predict the current collision risk between the vehicle and the target object based on the current motion information of the target object when the target object is present in the reflected image. The target object is used to represent an object that obstructs the vehicle's movement in the visual occlusion area. Based on the current collision risk, the module controls the vehicle to perform a collision avoidance warning.

[0030] Thirdly, this application provides a controller, including a storage module and a processing module. The storage module is used to store executable program code, and the processing module is used to call and run the executable program code from the storage module, causing the controller to execute the methods in the first aspect or any possible implementation of the first aspect.

[0031] Fourthly, this application provides a vehicle including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, causing the vehicle to perform the methods described in the first aspect or any possible implementation thereof.

[0032] Fifthly, this application provides a computer program product comprising: computer program code, which, when run on a computer, causes the computer to perform the method described in the first aspect or any possible implementation thereof.

[0033] Sixthly, this application provides a computer-readable storage medium storing computer program code that, when executed on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of a scenario for the vehicle collision avoidance warning method provided in the embodiments of this application; Figure 2 This is another scenario diagram of the vehicle collision avoidance warning method provided in the embodiments of this application; Figure 3 This is a schematic flowchart of a vehicle collision avoidance warning method provided in an embodiment of this application; Figure 4This is another scenario illustration of the vehicle collision avoidance warning method provided in the embodiments of this application; Figure 5 This is another schematic flowchart of a vehicle collision avoidance warning method provided in an embodiment of this application; Figure 6 This is a schematic diagram of the vehicle collision avoidance warning device provided in the embodiments of this application; Figure 7 This is a schematic diagram of the controller provided in an embodiment of this application; Figure 8 This is a schematic diagram of the vehicle structure provided in the embodiments of this application. Detailed Implementation

[0035] The technical solutions in this application will be clearly and thoroughly described below with reference to the accompanying drawings. In the description of the embodiments of this application, unless otherwise stated, " / " means "or," for example, A / B can mean A or B. "And / or" in the text is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of the embodiments of this application, "multiple" refers to two or more than two.

[0036] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0037] With the rapid development of vehicle intelligence technology, most vehicles are now equipped with intelligent collision avoidance systems. These systems provide collision warnings and control during driving to ensure vehicle safety. Specifically, while driving, vehicles can use onboard cameras to capture images of the area in front of them and detect obstacles. Upon detecting an obstacle, the system promptly issues a collision warning, effectively preventing a collision. However, if the area in front of the vehicle is obstructed, the camera may be unable to capture images, hindering obstacle detection and preventing timely collision warnings. This increases the risk of collision, making collisions more likely and reducing driving safety. The onboard cameras can include at least one of the following: panoramic cameras, front-view cameras, side-view cameras, and rear-view cameras.

[0038] Figure 1This is a schematic diagram of a scenario for the vehicle collision avoidance warning method provided in the embodiments of this application.

[0039] For example, such as Figure 1 As shown, Figure 1 This includes vehicle A 101 and pedestrian 102. Vehicle A 101 is equipped with an intelligent collision avoidance assistance system.

[0040] When vehicle A 101 is driving, it can use its own camera to collect images of the road ahead and perform obstacle detection on the collected images. When pedestrian 102 is detected, the intelligent collision avoidance system can provide collision warning and collision avoidance control for vehicle A 101 to avoid collision with pedestrian 102, thereby ensuring the driving safety of vehicle A 101 and the life safety of pedestrian 102.

[0041] But Figure 2 If vehicle B 103 is traveling in front of vehicle A 101 and obstructs vehicle A 101's view, specifically blocking the view of a pedestrian 102 in front of vehicle B 103, vehicle A 101's camera will be unable to capture an image of pedestrian 102. Consequently, vehicle A 101 will be unable to detect the pedestrian 102, and a collision is highly likely, threatening both vehicle A 101's driving safety and pedestrian 102's life. The blind spot created by vehicle B 103 obstructing vehicle A 101's view can be defined as the "visual obstruction area."

[0042] In related technologies, when vehicle A 101 has a visually obstructed area, vehicle A 101 requests an image of the visually obstructed area from vehicle B 103 to detect obstacles within the visually obstructed area using the image transmitted by vehicle B. However, this method relies on the premise that vehicle A 101 and vehicle B 103 can communicate. If vehicle A 101 and vehicle B 103 cannot communicate, then vehicle A 101 cannot obtain the image of the visually obstructed area transmitted by vehicle B, thus preventing vehicle A 101 from detecting pedestrian 102, and vehicle A 101 may still collide with pedestrian 102.

[0043] The following is combined with Figures 3 to 5 The vehicle collision avoidance warning method provided in the embodiments of this application will be described in detail.

[0044] Figure 3 This is a schematic flowchart of a vehicle collision avoidance warning method provided in an embodiment of this application. The method can be implemented by a vehicle (e.g., Figure 1 The vehicle 101 in the vehicle executes the command, or the controller in the vehicle executes the command.

[0045] For example, such as Figure 3 As shown, the method 300 includes the following implementation process: S310, when a visual obstruction area is detected in front of the vehicle, acquires reflected images of other target vehicles based on the visual obstruction area.

[0046] The reflected image represents the visual image reflected onto other target vehicles from a visually obstructed area, caused by external obstructions to the vehicle. Other target vehicles serve as the display medium for the reflected image, displaying the visual image reflected from the obstructed area. For example, other target vehicles might be oncoming vehicles. Figure 4 Vehicle C (104) in the list, or other target vehicles, are vehicles in front of the vehicle, such as... Figure 4 Vehicle D (105) in the diagram; this is because the vehicle body can act as a reflective surface, such as the paint, windows, and rearview mirror housings. Objects within the visually obstructed area (e.g., pedestrians, vehicles) reflect ambient light, and when this reflected light illuminates other target vehicles, reflected images of these objects are formed on those vehicles. Furthermore, the visually obstructed area is the blind spot of the vehicle's camera due to obstruction; the visual image includes at least one of the following: the outline, color, and motion information of each object within the visually obstructed area. Motion information may include at least one of the following: motion speed, direction, angle, and posture. Additionally, external obstructions are used to represent other obstructions besides the vehicle and any obstructions attached to it, i.e., external obstructions not existing on the vehicle, such as other vehicles, poles, trees, etc. For example, if leaves attached to the vehicle create a visually obstructed area, since the leaves are present on the vehicle, they are not considered external obstructions.

[0047] For example, when a vehicle is powered on, to avoid collisions, images of the area in front of the vehicle can be captured in real time for obstacle detection. However, if there is a visual obstruction area in front of the vehicle, the vehicle may not be able to directly capture images of that obstruction area using its own cameras, thus preventing obstacle detection in that area. In this case, to detect obstacles in the obstructed area, the visual image reflected from the obstructed area onto other target vehicles can be identified, and this reflected image can be used as the "reflected image on other target vehicles".

[0048] Optionally, when the vehicle is powered on, it can be detected whether the vehicle is moving. When the vehicle is moving, if there is a visual obstruction area in front of the vehicle, the reflected images of other target vehicles can be determined through the visual obstruction area in front of the vehicle. When the vehicle is not moving, it means that the vehicle is stationary. In this case, it can be continuously detected whether the vehicle is moving until it is detected that the vehicle is moving.

[0049] It should be understood that when determining the reflected images on other target vehicles, the vehicle can capture these reflected images using its own camera, thereby detecting obstacles in visually obstructed areas. There must be at least one other target vehicle. Furthermore, determining the reflected images on other target vehicles requires first confirming the existence of such vehicles; only if other target vehicles are present can their reflected images be determined. Currently, due to the increasing prevalence of vehicles and the large number of vehicles on the road, other target vehicles are generally present.

[0050] Optionally, when determining whether other target vehicles exist, the reflected image carriers (i.e., other target vehicles) corresponding to the visually obstructed area of ​​the vehicle can be identified. Specifically, the light path propagation direction of the visually obstructed area is determined, and other vehicles that can be captured by the vehicle's own camera in that light path propagation direction are identified as other target vehicles. That is, other target vehicles are those vehicles in the light path propagation direction that can be imaged by the vehicle's own camera; for example, oncoming vehicles in the opposite lane. Other vehicles not in the light path propagation direction, or those in the light path propagation direction but whose images cannot be captured by the vehicle's own camera, cannot be identified as other target vehicles.

[0051] S320, in response to the presence of a target object in the reflected image, predicts the current collision risk between the vehicle and the target object based on the current motion information of the target object.

[0052] The target object refers to an object that obstructs vehicle movement within a visually occluded area, such as a pedestrian or vehicle. There is at least one target object. Furthermore, the reflected image can be at least one consecutive frame, or multiple consecutive frames from different target vehicles, used to capture dynamic changes of the target object.

[0053] For example, after a vehicle captures reflected images of other target vehicles using its own camera, it can first analyze whether a target object exists in the reflected image. If a target object is found in the reflected image, it means there is an object obstructing the vehicle's movement in the obstructed area. To avoid a collision between the vehicle and the target object, in response to the presence of a target object in the reflected image, the current movement information of the target object (e.g., the aforementioned motion characteristics) can be extracted, and the current collision risk between the vehicle and the target object can be predicted using this information. If no target object is found in the reflected image, it means there is no object obstructing the vehicle's movement in the obstructed area, and the vehicle can drive normally in that area without a collision.

[0054] S330, based on the current collision risk, controls the vehicle to issue a collision avoidance warning.

[0055] For example, based on the current collision risk between the vehicle and a target object, the vehicle is controlled to issue a collision avoidance warning for that target object, thereby minimizing the possibility of a collision. The collision avoidance warning can be implemented through at least one of the following methods: voice, text, image, light, horn, etc.

[0056] In such Figure 3 In the method 300 shown, when a visual obstruction area is detected in front of the vehicle, in order to predict the collision risk between the vehicle and an object (i.e., a target object) obstructing the vehicle's movement within the visual obstruction area, reflected images on other target vehicles can be determined through the visual obstruction area. The current motion information of the target object in the reflected image is then used to predict the current collision risk between the vehicle and the target object, thereby controlling the vehicle to issue a collision avoidance warning. Compared to the problem of being unable to detect target objects within a visual obstruction area in front of the vehicle, thus preventing collision risk prediction, this embodiment utilizes the principle of optical reflection to obtain reflected images on other target vehicles through the visual obstruction area. This breaks through the physical boundaries of traditional visual perception, allowing indirect detection of target objects not within the vehicle's own visual range. Upon detection of the target object, the current motion information of the target object is used to predict the collision risk between the vehicle and the target object. This predictive collision risk allows for early collision avoidance warning, minimizing the possibility of collisions between the vehicle and the target object and effectively improving vehicle safety.

[0057] It should be noted that S310~S330 above is a simplified description of the vehicle collision avoidance warning method provided in the embodiments of this application. The following is a further explanation... Figure 3 The specific implementation methods shown in the embodiments are described in detail below: When executing S310, the above-mentioned acquisition of reflected images on other target vehicles based on visual occlusion areas includes: acquiring the current distance between other target vehicles and vehicles; determining the target reflected area on other target vehicles based on the visual occlusion area and the current distance; and determining the reflected image based on the visual information in the target reflected area.

[0058] Specifically, the size of the target reflection area is positively correlated with the size of the visual occlusion area, and there is a corresponding relationship between the two. Furthermore, the size of the target reflection area is positively correlated with the current distance, and there is a corresponding relationship between them. The size of the visual occlusion area can include at least one of the following range parameters: length, height, and area. The current distance between other target vehicles can be the Euclidean distance between other target vehicles.

[0059] For example, when other target vehicles are obtained, the current distance between the other target vehicles and the vehicle can be determined first, such as... Figure 4 The system uses the spacing 1 and spacing 2, as well as the area size of the visual occlusion region, to determine the target reflection areas on other target vehicles based on the current spacing between them and the area size of the visual occlusion region. Then, it uses the visual information displayed in the target reflection areas to determine the reflected images on other target vehicles.

[0060] Specifically, the optical path propagation area is determined by the visual obstruction area of ​​the vehicle; that is, the visual obstruction area is defined as the optical path propagation area. Furthermore, this optical path propagation area is affected by the current distance between other target vehicles. The smaller the current distance between other target vehicles, the smaller the optical path propagation area, resulting in a smaller reflection area from the visual obstruction area onto other target vehicles. Conversely, the larger the current distance between other target vehicles, the larger the optical path propagation area, resulting in a larger reflection area from the visual obstruction area onto other target vehicles. This reflection area from the visual obstruction area onto other target vehicles is defined as the target reflection area. This can be understood by analogy to the size of the light spot projected onto a wall by a flashlight: the closer the flashlight is to the wall, the smaller the light propagation area, resulting in a smaller light spot on the wall; conversely, the farther the flashlight is from the wall, the larger the light propagation area, resulting in a larger light spot on the wall.

[0061] In this embodiment, when determining the target reflection area on other target vehicles, the determination is made by combining the visual occlusion area of ​​the vehicle with the current distance between the other target vehicles. This takes into account the influence of the current distance between the other target vehicles on the target reflection area, avoiding potential biases that may occur when determining the target reflection area solely based on the visual occlusion area of ​​the vehicle, thus improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for the vehicle and the target object, thereby further preventing collisions between the vehicle and the target object.

[0062] In one implementation, determining the target reflection area on other target vehicles based on the visual occlusion area and the current distance includes: determining the initial reflection area on other target vehicles based on the visual occlusion area and the current distance; and adjusting the initial reflection area based on the current road slope of other target vehicles to obtain the target reflection area.

[0063] Specifically, the size of the visually occluded area is positively correlated with the size of the initial reflective area, and there is a corresponding relationship between the size of the visually occluded area and the size of the initial reflective area. Furthermore, the current spacing is positively correlated with the size of the initial reflective area, and there is a corresponding relationship between the current spacing and the size of the initial reflective area.

[0064] For example, when determining the target reflection area, instead of directly determining the target reflection area through the current distance between other target vehicles and the area size of the visual occlusion area, the initial reflection area on other target vehicles can be determined first through the current distance between other target vehicles and the area size of the visual occlusion area. This initial reflection area has not been adjusted by the current road slope of other target vehicles.

[0065] Then, the current road slope of other target vehicles is obtained, and the area size and / or area position of the initial reflection area are adjusted based on the current road slope of other target vehicles to obtain the adjusted reflection area, and the adjusted reflection area is determined as the target reflection area.

[0066] The vehicle can acquire real-time road slope data of other target vehicles using an onboard slope sensor (e.g., a gyroscope). Alternatively, a vehicle-mounted positioning device can acquire real-time vehicle location data, and then use this location data to determine the road slope of that location in map data, i.e., the road slope of other target vehicles. The map data can include any of the following: high-precision maps, navigation maps, etc. The positioning device can include at least one of the following: inertial measurement unit (IMU), wheel odometer, Global Positioning System (GPS), etc. It should be understood that the onboard slope sensor and positioning device are common vehicle configurations, and no additional impact is required to implement the embodiments of this application, thus reducing the implementation cost of the embodiments of this application.

[0067] Optionally, the target reflection area on other target vehicles can be determined based on the visual occlusion area and the current distance, as well as the current road slope of other target vehicles.

[0068] In this embodiment, when determining the target reflection area on other target vehicles, the current road slope of the other target vehicles is also considered in addition to the vehicle's visual occlusion area and the current distance between other target vehicles. This avoids potential deviations when determining the target reflection area based solely on the vehicle's visual occlusion area and the current distance between other target vehicles, further improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for vehicles and target objects, thereby further preventing collisions between vehicles and target objects.

[0069] Furthermore, the above-mentioned adjustment of the area size of the initial reflection area based on the current road slope of other target vehicles to obtain the target reflection area includes: determining the target adjustment amount based on the current road slope; and adjusting the initial reflection area based on the target adjustment amount to obtain the target reflection area.

[0070] The target adjustment amount is positively correlated with the current road slope, and there is a corresponding relationship between the adjustment amount and the road slope. Furthermore, the target adjustment amount includes a target scaling amount that scales the initial reflection area's size, and / or a target adjustment distance that moves the initial reflection area's location.

[0071] For example, firstly, based on the current road slope of other target vehicles, the adjustment amount corresponding to the current road slope (which can be called the "target adjustment amount") is determined according to the correspondence between the adjustment amount and the road slope. Then, the area size and / or area position of the initial reflection area are adjusted using this target adjustment amount to obtain the adjusted reflection area, which is then identified as the target reflection area. Specifically, the area size of the initial reflection area is adjusted using a target scaling amount to obtain the adjusted reflection area, which is then identified as the target reflection area. And / or, the area position of the initial reflection area is adjusted using a target adjustment distance to obtain the adjusted reflection area, which is then identified as the target reflection area.

[0072] For example, when the current road slope of other target vehicles is uphill, the location of the initial reflection area can be moved upwards towards the vehicle body by the target adjustment distance, and / or the size of the initial reflection area can be scaled by the target scaling amount.

[0073] For example, when the current road slope of other target vehicles is downhill, the location of the initial reflection area can be moved downwards from the vehicle body by a target adjustment distance, and / or the size of the initial reflection area can be scaled by a target scaling amount.

[0074] In this embodiment, the reflection areas determined by the visual occlusion area of ​​passing vehicles, the current distance between other target vehicles, and the current road slope of other target vehicles are adjusted to obtain the target reflection areas on other target vehicles, further improving the accuracy of the target reflection areas. Furthermore, with more accurate target reflection areas, more accurate collision risk prediction can be made for vehicles and target objects, thereby further avoiding collisions between vehicles and target objects.

[0075] Optionally, the above-mentioned adjustment of the region size of the initial reflection region based on the target adjustment amount to obtain the target reflection region includes: adjusting the initial reflection region based on the target adjustment amount to obtain the adjusted reflection region; and determining the region of interest in the adjusted reflection region as the target reflection region.

[0076] Specifically, the probability of visual information existing in a visually occluded area within the region of interest (ROI) is greater than the probability of visual information existing in a visually occluded area within a non-ROI region in the adjusted reflective region. The non-ROI region refers to the area within the adjusted reflective region excluding the ROI. Furthermore, the probability of visual information existing in a visually occluded area within the ROI is greater than or equal to a preset probability. The preset probability represents the minimum probability of visual information existing in a visually occluded area within the ROI, such as 80% or 90%, and can be preset; this embodiment does not limit this.

[0077] For example, when the adjusted reflection area is obtained, instead of directly defining the adjusted reflection area as the target reflection area, a region of interest (ROI) can be defined within the adjusted reflection area, and this ROI can be defined as the target reflection area. ROI represents a specific area that needs to be focused on.

[0078] Optionally, the region of interest (ROI) is determined using the current motion information of the target object. Specifically, the target object's motion intention is determined using its current motion information. The potential region of the target object within the adjusted reflection area is predicted based on its motion intention, and this potential region is then identified as the ROI. For example, if the target object's motion intention is to move forward, the display area in the adjusted reflection area when the target object moves forward can be identified as a potential region, and this potential region is then identified as the ROI. Similarly, if the target object's motion intention is to stand and observe, the display area in the adjusted reflection area when the target object stands and observes can be identified as a potential region, and this potential region is then identified as the ROI. Finally, if the target object's motion intention is to move backward, the display area in the adjusted reflection area when the target object moves backward can be identified as a potential region, and this potential region is then identified as the ROI.

[0079] Optionally, upon obtaining the adjusted reflection area, the center point of the adjusted reflection area can be calculated using image coordinates. A central region including the center point can then be calculated using this center point, and this central region can be defined as the target reflection area. This allows collision risk prediction to focus on the central region, thereby reducing the computational load of collision risk prediction. Specifically, with this center point as the center, the product of the length and width of the adjusted reflection area and a preset multiple is used to determine the length and width of the central region. The preset multiple is a positive number less than 1.

[0080] In this embodiment, the reflection area determined by the adjustment amount based on the current road slope of other target vehicles is adjusted based on the visual occlusion area of ​​the vehicle, the current distance between other target vehicles, and the distance between vehicles to obtain the adjusted reflection area, thus improving the accuracy of the adjusted reflection area. Furthermore, based on the more accurate adjusted reflection area, a more accurate target reflection area can be obtained. Additionally, defining the region of interest (ROI) within the adjusted reflection area as the target reflection area avoids interference from non-ROI regions, further improving the accuracy of the target reflection area. Moreover, by selecting the ROI, the amount of data for subsequent processing can be effectively reduced. It is not necessary to perform full calculation on the adjusted reflection area; only a partial calculation of the ROI is required, allowing collision risk prediction to focus on the ROI, thereby reducing the computational load of collision risk prediction.

[0081] In one implementation, the above-mentioned acquisition of reflected images on other target vehicles based on visual occlusion areas includes: determining target reflected areas on other target vehicles based on the geometric dimensions of the visual occlusion areas; and determining reflected images based on visual information in the target reflected areas.

[0082] Specifically, the geometric dimensions of the target reflection area are positively correlated with the geometric dimensions of the visual occlusion area. The geometric dimensions represent at least one of the shape parameters of the area's own outline, such as length, width, and height; for example, the height of the reflection area and the height of the visual occlusion area. It should be understood that geometric dimensions and area dimensions are not the same; area dimensions represent the area coverage parameter, and the two are not identical. Furthermore, there is a corresponding relationship between the geometric dimensions of the visual occlusion area and the target reflection area.

[0083] For example, the target reflection area corresponding to the geometric dimensions of the line-of-sight obstruction area is determined by using the geometric dimensions of the obstruction area. The target reflection area corresponding to the height of the obstruction area is determined by using the height of the obstruction area.

[0084] For example, when the height of the obstructed view area is less than a first preset height, the headlight area of ​​other target vehicles is designated as the target reflection area. When the height of the obstructed view area is greater than or equal to the first preset height and less than or equal to a second preset height, the front bumper area of ​​other target vehicles is designated as the target reflection area. When the height of the obstructed view area is greater than the second preset height, the windshield area of ​​other target vehicles is designated as the target reflection area. The first preset height is less than the second preset height. The first and second preset heights can be obtained through actual vehicle testing and pre-stored in the vehicle's storage unit or in a cloud server connected to the vehicle for easy retrieval. For example, the first preset height is 1.8m and the second preset height is 2.8m.

[0085] Optionally, the target reflection area is determined based on the geometry of the visual obstruction area and the current distances between other target vehicles. When determining the target reflection area on other target vehicles, the current distances between other target vehicles are considered in conjunction with the geometry of the visual obstruction area. This takes into account the influence of the current distances between other target vehicles on the target reflection area, avoiding potential biases that may occur when determining the target reflection area solely based on the geometry of the visual obstruction area, thus improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, more accurate collision risk prediction can be performed on vehicles and target objects, thereby further preventing collisions between vehicles and target objects.

[0086] For example, the reflection area corresponding to the geometric dimensions of the obstructed view is determined by using the geometric dimensions of the obstructed view area. Then, the reflection area corresponding to the geometric dimensions of the obstructed view area is adjusted based on the current distances between other target vehicles, resulting in an adjusted reflection area. The region of interest within the adjusted reflection area is then determined as the target reflection area. The size of the target reflection area is positively correlated with the current distance.

[0087] Specifically, a first adjustment amount is determined based on the current distances between other target vehicles and other vehicles. This first adjustment amount is then used to adjust the reflection area corresponding to the geometric dimensions of the obstructed view, resulting in the adjusted reflection area. The first adjustment amount is positively correlated with the current distances between other target vehicles and other vehicles, and a corresponding relationship exists between the first adjustment amount and the current distances between other target vehicles and other vehicles.

[0088] In this embodiment, when determining the target reflection area on other target vehicles, the geometric dimensions of the visual occlusion area are used for determination. This ensures a stable mapping relationship between the determined target reflection area and the geometric dimensions of the visual occlusion area, thereby improving the accuracy of the target reflection area. Furthermore, with a more accurate target reflection area, a more accurate collision risk prediction can be made for vehicles and target objects, thereby further avoiding collisions between vehicles and target objects.

[0089] When executing S320, the above-mentioned prediction of the current collision risk between the vehicle and the target object based on the target object's current motion information includes: determining the target object's current motion intention based on the target object's current motion information; determining the vehicle's current driving intention based on the vehicle's current operating information; and predicting the current collision risk based on the target object's current motion intention and the vehicle's current driving intention.

[0090] For example, the current motion intention of the target object can be determined by its current speed and / or current direction of motion. Furthermore, to improve the accuracy of collision risk prediction when determining the current collision risk between the vehicle and the target object, the vehicle's current operating information can be obtained, and its current driving intention can be determined using this information. The current collision risk can then be predicted by combining the target object's current motion intention and the vehicle's current driving intention.

[0091] Optionally, the current motion intention of the target object can be determined by the current motion speed of the target object and / or the rate of change of the current direction of movement of the target object.

[0092] When the target object's current speed is less than or equal to the preset speed, it indicates that the target object is essentially stationary, and its current intention can be predicted as standing still and observing. When the target object's current speed is greater than the preset speed, and the rate of change of its current direction of movement is less than or equal to the preset rate of change, it indicates that the target object is moving rapidly without changing its direction of movement, and its current intention can be predicted as moving forward. When the rate of change of the target object's current direction of movement is greater than the preset rate of change, it indicates that the target object is moving with a change in direction of movement, and its current intention can be predicted as moving backward (which can be called "mid-journey turnaround").

[0093] The target object's current movement intention can be any of the following: moving forward, moving backward, or standing still and observing. Furthermore, there is a correspondence between the target object's current movement intention and the current collision risk between the vehicle and the target object.

[0094] For example, if the target object's current intention is to move forward, there is a high chance of it "suddenly appearing" (or "ghosting"), making a collision with the vehicle highly probable. In this case, the risk of a collision between the vehicle and the target object is extremely high. Conversely, if the target object's current intention is to stand still and observe, there is less chance of it "suddenly appearing," and a collision with the vehicle is generally unlikely. Therefore, the risk of a collision between the vehicle and the target object is low. And if the target object's current intention is to move backward, there is even less chance of it "suddenly appearing," and a collision with the vehicle is generally unlikely. Therefore, the risk of a collision between the vehicle and the target object is extremely low.

[0095] The preset speed represents the maximum speed of the target object when it is stationary, such as 0.1 m / s or 0.2 m / s. The preset rate of change represents the maximum rate of change when the target object's direction of movement has not changed, such as 10% or 12%.

[0096] The vehicle's current operating information may include at least one of the following: accelerator pedal opening, brake pedal opening, steering wheel angle, and vehicle speed. This information can be collected by various sensors within the vehicle, such as wheel speed sensors for vehicle speed, accelerator pedal position sensors for accelerator pedal opening, brake pedal position sensors for brake pedal opening, and steering wheel angle sensors for steering wheel angle.

[0097] When the brake pedal opening increases, it indicates that the vehicle is decelerating, so the current driving intention can be predicted as braking. When the accelerator pedal opening increases, it indicates that the vehicle is accelerating, so the current driving intention can be predicted as accelerating forward. When the steering wheel angle is turned to the left, it indicates that the vehicle is about to turn left, so the current driving intention can be predicted as turning left forward.

[0098] The vehicle's current driving intention can be any of the following: deceleration / braking, acceleration, or turning left. Furthermore, there is a correspondence between the target object's current motion intention, the vehicle's current driving intention, and the current collision risk between the vehicle and the target object.

[0099] For example, if the target object's current intention is to move forward, and the vehicle's current intention is to accelerate forward, the "ghost peek" will appear more quickly, and the target object is highly likely to collide with the vehicle. In this case, the risk of a collision between the vehicle and the target object is extremely high. Conversely, if the target object's current intention is to move forward, and the vehicle's current intention is to decelerate or brake, the "ghost peek" will appear with a delay, and the target object may still collide with the vehicle. In this case, the risk of a collision between the vehicle and the target object is relatively high. Finally, if the target object's current intention is to move forward, and the vehicle's current intention is to turn left, a "ghost peek" is less likely to occur, and a collision is generally unlikely. Therefore, the risk of a collision between the vehicle and the target object is relatively low.

[0100] For example, if the target's current intention is to stop and observe, and the vehicle's current intention is to accelerate forward, this can cause a time delay in the appearance of a "ghost peek" (emerging suddenly from behind a vehicle). The target may still collide with the vehicle, in which case the current collision risk between the vehicle and the target is high. Conversely, if the target's current intention is to stop and observe, and the vehicle's current intention is to decelerate and brake, a "ghost peek" is unlikely, and a collision is generally unlikely; therefore, the current collision risk between the vehicle and the target is low. Finally, if the target's current intention is to stop and observe, and the vehicle's current intention is to turn left, a "ghost peek" is even less likely, and a collision is generally unlikely; therefore, the current collision risk between the vehicle and the target is extremely low.

[0101] For example, if the target object's current intention is to move backward, and the vehicle's current intention is to accelerate forward, a "ghost peek" situation is unlikely, and a collision is generally unlikely. Therefore, the current collision risk between the vehicle and the target object is low. Similarly, if the target object's current intention is to move backward, and the vehicle's current intention is to decelerate or brake, a "ghost peek" situation is also unlikely, and a collision is generally unlikely. Therefore, the current collision risk between the vehicle and the target object is low. And if the target object's current intention is to move backward, and the vehicle's current intention is to turn left forward, a "ghost peek" situation is even less likely, and a collision is generally unlikely. Therefore, the current collision risk between the vehicle and the target object is extremely low.

[0102] Optionally, based on the current motion information of the target object, the current motion intention of the target object is determined; based on the current motion intention of the target object, the current collision risk is predicted.

[0103] For example, the current motion intention of the target object can be determined by its current speed and / or current direction of motion. Then, based on the target object's current motion intention, the current collision risk between the vehicle and the target object can be predicted. Determining the target object's current motion intention through its current motion information allows for the prediction of its future motion trend, thus enabling the early prediction of the current collision risk between the vehicle and the target object. This provides sufficient warning time for the vehicle, thereby improving the timeliness and reliability of collision warnings.

[0104] In this embodiment, the collision risk between the vehicle and the target object is predicted by combining the target object's current motion intention with the vehicle's current driving intention. This approach considers the impact of the vehicle's current driving intention on the collision risk prediction result, avoiding potential biases that might arise from predicting collision risk solely based on the target object's current motion intention, thus improving the accuracy of the collision risk prediction. Furthermore, with a more accurate collision risk prediction result, collisions between the vehicle and the target object are further avoided. Additionally, by determining the vehicle's current driving intention through its current operating information, the future movement trend of the vehicle can be predicted. This allows for advance prediction of the current collision risk between the vehicle and the target object based on both the target object's current motion intention and the vehicle's current driving intention, providing sufficient warning time for the vehicle and further improving the timeliness and reliability of the collision warning.

[0105] Furthermore, the above-mentioned prediction of the current collision risk based on the current motion intention of the target object and the current driving intention of the vehicle includes: when there is only one target object, predicting the current collision risk based on the current motion intention of the one target object and the current driving intention of the vehicle; when there are multiple target objects, determining a first target object based on the current distance between each target object and the vehicle; predicting the current collision risk based on the current motion intention of the first target object and the current driving intention of the vehicle; or, when there are multiple target objects, determining the target collision risk between the vehicle and each target object based on the current distance between each target object and the vehicle, the current motion intention of each target object, and the current driving intention of the vehicle; and performing risk fusion on multiple target collision risks to obtain the current collision risk.

[0106] Among these, the first target object is defined as the one with the smallest current distance to the vehicle. The current distance between the target object and the vehicle can be the Euclidean distance. Furthermore, the risk fusion method can be any of the following: weighted fusion, Bayesian fusion, machine learning fusion, etc.

[0107] Optionally, the current collision risk between the vehicle and the target object can be predicted based on the target object's current motion intention, the vehicle's current driving intention, and the number of target objects.

[0108] For example, when there is only one target object, the current collision risk between the vehicle and the target object can be predicted directly by the current motion intention of that one target object and the current driving intention of the vehicle.

[0109] Optionally, the initial collision risk between the vehicle and the target object is first determined by the current motion intention of the target object and the current driving intention of the vehicle. Then, the weight of the target object is determined by the current distance between the target object and the vehicle. Next, the initial collision risk between the vehicle and the target object, and the weight of the target object are weighted to obtain the target collision risk between the vehicle and the target object. Specifically, the product of the initial collision risk between the vehicle and the target object and the weight of the target object is determined as the target collision risk between the vehicle and the target object.

[0110] For example, when there are multiple target objects, the current distance between each target object and the vehicle can be obtained first. These distances are then sorted, and the target object with the shortest current distance to the vehicle (referred to as the "first target object") is selected. The current motion intention of this first target object and the current driving intention of the vehicle are then used to predict the current collision risk between the vehicle and the first target object. This is because the first target object, with the shortest current distance, will be the first to collide with the vehicle. Therefore, prioritizing the collision risk between the first target object and the vehicle allows for the prediction of the current collision risk based on the current motion intention of the first target object and the current driving intention of the vehicle.

[0111] For example, the target objects are target object A, target object B, and target object C. The current distance between target object A and the vehicle is 3m, the current distance between target object B and the vehicle is 2m, and the current distance between target object C and the vehicle is 5m. Since 2m < 3m < 5m, target object B can be identified as the first target object.

[0112] Optionally, the initial collision risk between the vehicle and the first target object is first determined by the current movement intention of the first target object and the current driving intention of the vehicle. Then, the weight of the first target object is determined by the current distance between the first target object and the vehicle. Next, the initial collision risk between the vehicle and the first target object and the weight of the first target object are weighted to obtain the target collision risk between the vehicle and the first target object. Specifically, the product of the initial collision risk between the vehicle and the first target object and the weight of the first target object is determined as the target collision risk between the vehicle and the first target object.

[0113] For example, when there are multiple target objects, the current collision risk between each target object and the vehicle can be determined first, and then the current collision risk between each target object and the vehicle can be fused to obtain the current collision risk between the vehicle and multiple target objects as a whole.

[0114] Specifically, based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle, the target collision risk between the vehicle and each target object is determined.

[0115] For example, given target objects A, B, and C, the target collision risk A between the vehicle and target object A can be determined using the current distance between target object A and the vehicle, the current movement intention of target object A, and the current driving intention of the vehicle. Similarly, the target collision risk B between the vehicle and target object B can be determined using the current distance between target object B and the vehicle, the current movement intention of target object B, and the current driving intention of the vehicle. Likewise, the target collision risk C between the vehicle and target object C can be determined using the current distance between target object C and the vehicle, the current movement intention of target object C, and the current driving intention of the vehicle. Then, target collision risks A, B, and C are fused to obtain the overall current collision risk between the vehicle and target objects A, B, and C.

[0116] In this embodiment, selecting different collision risk prediction methods based on the number of target objects makes collision risk prediction between vehicles and target objects more flexible and more consistent with the actual situation of the target objects, further improving the accuracy of collision risk prediction results. Consequently, based on more accurate collision risk prediction results, collisions between vehicles and target objects can be further avoided.

[0117] Furthermore, the above-mentioned determination of the target collision risk between the vehicle and each target object based on the current distance between each target object and the vehicle, the current motion intention of each target object, and the current driving intention of the vehicle includes: determining the initial collision risk between the vehicle and each target object based on the current motion intention of each target object and the current driving intention of the vehicle; determining the weight of each target object based on the current distance between each target object and the vehicle; and weighting the initial collision risk between the vehicle and each target object and the weight of each target object to obtain the target collision risk between the vehicle and each target object.

[0118] The weight of each target object is negatively correlated with the current distance between it and the vehicle; that is, the shorter the current distance, the greater the weight; and the longer the current distance, the smaller the weight. There is a direct correlation between the weight of each target object and its current distance to the vehicle.

[0119] For example, the initial collision risk between the vehicle and each target object is first determined by the current motion intention of each target object and the current driving intention of the vehicle. Then, the weight of each target object is determined by the current distance between each target object and the vehicle.

[0120] Furthermore, the initial collision risk between the vehicle and each target object is weighted and the weights of each target object are averaged to obtain the target collision risk between the vehicle and each target object. Specifically, the product of the initial collision risk between the vehicle and each target object and the weights of each target object is determined as the target collision risk between the vehicle and each target object.

[0121] For example, the target objects are target object A, target object B, and target object C. The current distance between target object A and the vehicle is 3m, the current distance between target object B and the vehicle is 2m, and the current distance between target object C and the vehicle is 5m. The weights for target object A and target object B are 0.5, 0.6, and 0.8, respectively. The target collision risk for target object A is: initial collision risk of target object A × 0.5. The target collision risk for target object B is: initial collision risk of target object B × 0.6. The target collision risk for target object C is: initial collision risk of target object C × 0.8.

[0122] It should be understood that there is a corresponding relationship between the current distance between each target object and the vehicle and its weight. Furthermore, the initial collision risk is an unweighted collision risk.

[0123] In this embodiment, when determining the collision risk between a vehicle and a target object, the impact of the current distance between the target object and the vehicle on the collision risk is considered. The collision risk is adjusted using a weight determined by the current distance between the target object and the vehicle, which improves the accuracy of the collision risk prediction results. Furthermore, based on more accurate collision risk prediction results, collisions between the vehicle and the target object can be further avoided.

[0124] Figure 5 This is another schematic flowchart illustrating a vehicle collision avoidance warning method provided in an embodiment of this application. The method can be implemented by a vehicle (e.g., Figure 1 The vehicle 101 in the vehicle executes the command, or the controller in the vehicle executes the command.

[0125] For example, such as Figure 5 As shown, the method 500 includes the following implementation process: S510: Check if the vehicle is moving. If yes, proceed to S520; otherwise, proceed to S510.

[0126] For example, as vehicles become more widespread, traffic conditions are becoming increasingly complex, making vehicle collisions more likely. In complex road scenarios such as turning right at a red light, a visual obstruction area in the left lane, and pedestrians crossing at a crosswalk, the vehicle's own cameras cannot directly capture pedestrians within the visual obstruction area. This prevents the prediction of collision risks between pedestrians and vehicles, hindering early collision avoidance warnings and making collisions extremely likely. Therefore, to minimize the risk of collisions, the system can detect whether the vehicle is moving when it is powered on. If the vehicle is moving, S520 can be executed. If the vehicle is not moving, S510 can continue until vehicle movement is detected.

[0127] S520: Detect if there is a visual obstruction area in front of the vehicle. If yes, proceed to S530; otherwise, proceed to S520.

[0128] For example, by analyzing images captured by the vehicle itself, it is determined whether there is a blind spot in front of the vehicle. If a blind spot is detected, it indicates that there is an obstruction in front of the vehicle, and the area where the blind spot is located can be defined as the visual obstruction area, i.e., there is a visual obstruction area in front of the vehicle, and S530 can be executed. If no blind spot is detected, it indicates that there is no obstruction in front of the vehicle, i.e., there is no visual obstruction area in front of the vehicle, and S520 can continue to be executed until a visual obstruction area is detected in front of the vehicle.

[0129] S530: Based on the visual occlusion area, determine whether there are reflected images from other target vehicles. If yes, proceed to S540; otherwise, proceed to S580.

[0130] For example, the presence of reflected images from other target vehicles can be determined first by examining the visual occlusion area of ​​the vehicle. If reflected images from other target vehicles exist, i.e., if valid reflected images exist, S540 can be executed. If reflected images from other target vehicles do not exist, i.e., if no valid reflected images exist, S580 can be executed.

[0131] Optionally, S580 can also be executed even if a valid reflected image exists but its signal-to-noise ratio (SNR) is poor. For example, if the SNR of the valid image is less than or equal to a preset SNR, it indicates that the SNR of the valid image is poor, and S580 can be executed. The preset SNR can represent the highest SNR that cannot be identified due to the image quality being too low, such as 15dB or 20dB.

[0132] S540: Detect whether a target object exists in the reflected image. If yes, proceed to S550; otherwise, proceed to S540.

[0133] For example, a lightweight object detection neural network model (You Only Look Once version 5small, YOLOv5s) is used to detect objects (e.g., pedestrians, other vehicles) in the reflected image, and when a target object is detected, S550 is executed. If no target object is detected, S540 can be executed until the target object is detected in the reflected image. YOLOv5s can be configured in vehicles, such as the NVIDIA Xavier automotive embedded platform.

[0134] Optionally, upon detecting a target object, a distortion correction method is used to correct the distortion of the target object to obtain an undistorted target object, thereby ensuring the accuracy of the visual information (also known as "visual features") of the target object. The distortion correction method can include any of the following: Contrast Limited Adaptive Histogram Equalization (CLAHE), Polynomial Distortion Correction, Chessboard Calibration, or Camera Calibration Correction. CLAHE is used to enhance local image contrast.

[0135] Optionally, when obtaining the reflected image, an optical filter can be used to filter the reflected image to select the high-reflectivity band in the visible spectrum. The presence of a target object in the reflected image can then be detected using this high-reflectivity band. Detecting the presence of a target object in the reflected image using the high-reflectivity band can suppress interference from ambient light, improve the signal-to-noise ratio of the reflected image, and thus improve the accuracy of target object identification. Ambient light interference can be at least one of direct sunlight, vehicle headlight glare, and ambient stray light. The optical filter can be a filter used to select the high-reflectivity band in the visible spectrum, such as a 520nm-560nm (nanometer) narrowband optical filter, enhancing the contrast between the target object and the background.

[0136] S550 determines the current motion intention of the target object based on the target object's current motion information.

[0137] Optionally, the type of the target object can be identified, and the current movement intention of the target object can be determined based on its current motion information and type. Specifically, when the target object is a pedestrian, the current torso swing frequency of the target object can be analyzed through its current motion information. Then, based on the target object's current motion speed and / or current motion direction and the current torso swing frequency, the target object's current movement intention can be determined.

[0138] For example, when the target object is a pedestrian, the fluctuation of the pedestrian's shoulder keypoint angle can be analyzed, and the current torso swing frequency of the target object can be determined through this fluctuation. The current motion information and the current torso swing frequency of the target object are input into an intent analysis model (which can be called a "composite confidence model"). The intent analysis model is used to analyze the pedestrian's motion intent (i.e., the target object's current motion intent) corresponding to the current motion information and the current torso swing frequency. The shoulder keypoints can include at least one of the pedestrian's neck, left shoulder, and right shoulder. The intent analysis model is a pre-trained model. The intent analysis model can be at least one of Support Vector Machine (SVM), Decision Tree (DT), Hidden Markov Model (HMM), etc.

[0139] S560 predicts the current collision risk between the vehicle and the target object based on the target object's current motion intention.

[0140] Optionally, the remaining time to collision (TTC) between the vehicle and the target object can be calculated using the current distance between the target object and the vehicle, the current speed of the vehicle, and the current lateral deflection angle between the target object and the vehicle. TTC can be used to assess the urgency of a collision between the vehicle and the target object. TTC is negatively correlated with the urgency of a collision between the vehicle and the target object, and there is a corresponding relationship between TTC and the urgency of a collision between the vehicle and the target object.

[0141]

[0142] In formula (1), D represents the current distance between the target object and the vehicle, v represents the current speed of the vehicle, and θ represents the current lateral deflection angle between the target object and the vehicle.

[0143] S570, based on the current collision risk, controls the vehicle to issue a collision avoidance warning.

[0144] Optionally, the current collision level between the vehicle and the target object is determined based on the current collision risk. Then, the collision avoidance warning method for the target object is determined based on the current collision level, so as to control the vehicle to use the collision avoidance warning method to warn of the target object, thereby achieving a graded response for collision avoidance warnings. Specifically, the current collision risk and the current collision level between the vehicle and the target object are positively correlated, and there is a corresponding relationship between them. Furthermore, the current collision level between the vehicle and the target object is positively correlated with the type and intensity of the collision avoidance warning method; there is also a corresponding relationship between them. The collision level can be a classification system of extremely high, relatively high, moderate, and low. The collision avoidance warning method can be at least one of the following: yellow warning, orange warning, red warning, audible and visual warning, and braking preparation. For example, when the collision level is extremely high, the corresponding collision avoidance warning method is red warning, hazard horn, and automatic emergency braking (AED) pre-charging preparation. When the collision level is high, the corresponding collision warning method is an orange warning sign, a single horn sound, and automatic braking. When the collision level is low, the corresponding collision warning method is a yellow warning sign. An AED (Automated External Defibrillator) can be considered a vehicle active safety feature, capable of actively controlling the vehicle's braking.

[0145] Optionally, to improve the accuracy of collision risk prediction, multiple consecutive frames of reflection images can be collected to predict the collision risk between the target object and the vehicle in each reflection image. It can be determined whether the collision risk between the target object and the vehicle in each reflection image is consistent. If they are consistent, it indicates that the accuracy of the collision risk prediction is high, which can avoid misjudgment of single frame data. In this way, the predicted collision risk can be used to control the vehicle for collision avoidance warning.

[0146] S580, waiting for real-time images collected by the vehicle itself.

[0147] For example, when there are no reflected images of other target vehicles, the target object cannot be detected using reflected images. Therefore, the vehicle can only wait for its own real-time images to detect the target object. However, since the time when the target object is detected by the real-time images is generally later than the time when the target object is detected by the reflected images, this will cause a certain delay in the vehicle's collision avoidance warning, increasing the risk of collision between the vehicle and the target object. This is because the vehicle can only detect the target object when it enters its field of vision, while reflected images, using the principle of optical reflection, allow the vehicle to indirectly see the target object even when it has not entered its field of vision.

[0148] S590: Detect whether a target object exists in the real-time image. If yes, proceed to S550; otherwise, proceed to S590.

[0149] For example, when a target object is detected in the real-time image captured by the vehicle itself, S550 can be executed. If the target object is not found in the real-time image, S590 can be executed until the target object is detected.

[0150] It should be noted that, Figure 5 All steps are in Figures 3 to 4 The corresponding embodiments are described in detail, and will not be repeated here.

[0151] In summary, in this embodiment, when there is a visually obstructed area in front of the vehicle, obstacle (i.e., target object) detection can be achieved through reflected images from other target vehicles without relying on roadside equipment around the vehicle, thus ensuring the feasibility of collision risk prediction between the vehicle and the target object. This is accomplished collaboratively by the vehicle's camera, Controller Area Network (CAN), and processing unit, without increasing the vehicle's hardware costs, thus reducing processing costs. It should be understood that while roadside equipment has non-line-of-sight detection capabilities to detect visually obstructed areas, it relies on road infrastructure deployment, has low coverage, and cannot detect visually obstructed areas of vehicles in all locations, limiting its application scenarios. Although infrared thermal imaging cameras in vehicles can detect target objects in visually obstructed areas, infrared thermal imaging is susceptible to ambient temperature interference, has limited accuracy, and may confuse the thermal signals of target objects with those of other vehicles, leading to detection errors. Furthermore, while radar devices in vehicles can also detect target objects in visually obstructed areas, radar devices are susceptible to electromagnetic interference, have limited detection range, and cannot achieve accurate target object detection. Furthermore, while pedestrian detection can be achieved via Bluetooth signals from pedestrian mobile terminals, it cannot detect obstacles when the pedestrian is not carrying a mobile terminal or when the target is an animal. Moreover, even if a pedestrian is detected, the pedestrian's movement intention cannot be identified, which has significant limitations.

[0152] It should be understood that the correspondences involved in the embodiments of this application can all be obtained through real vehicle testing and pre-stored in the vehicle's storage unit, or stored in a cloud server that communicates with the vehicle, so that the vehicle can retrieve them at any time.

[0153] It should be understood that the above examples are provided to help those skilled in the art understand the embodiments of this application, and are not intended to limit the embodiments of this application to the specific values ​​or scenarios exemplified. Those skilled in the art can obviously make various equivalent modifications or variations based on the above examples, and such modifications or variations also fall within the scope of the embodiments of this application.

[0154] The above text combined Figures 1 to 5 The vehicle collision avoidance warning method provided in the embodiments of this application is described in detail below; the following will be combined with Figure 6 and Figure 7 The apparatus embodiments of this application are described in detail below. It should be understood that the apparatus in the embodiments of this application can perform the various methods described in the foregoing embodiments of this application, that is, the specific working processes of the various products described below can be referred to the corresponding processes in the foregoing method embodiments.

[0155] Figure 6 This is a schematic diagram of the vehicle collision avoidance warning device provided in the embodiments of this application.

[0156] For example, such as Figure 6 As shown, the device 600 includes: The acquisition module 610 is used to acquire reflected images on other target vehicles based on the visual occlusion area when a visual occlusion area is detected in front of the vehicle. The reflected images are used to represent the visual images reflected from the visual occlusion area to other target vehicles. The visual occlusion area is caused by external obstructions of the vehicle. The processing module 620 is used to predict the current collision risk between the vehicle and the target object based on the current motion information of the target object when the target object is present in the reflected image, wherein the target object is used to represent an object that obstructs the vehicle's movement in the visual occlusion area; and to control the vehicle to perform a collision avoidance warning based on the current collision risk.

[0157] In one possible implementation, the acquisition module 610 is specifically used for: Get the current distances between other target vehicles and between vehicles; Based on the visual occlusion area and the current distance, the target reflection area on other target vehicles is determined, wherein the area size of the target reflection area is positively correlated with the area size of the visual occlusion area and the area size of the target reflection area is positively correlated with the current distance; The reflected image is determined based on visual information in the target's reflective area.

[0158] In one possible implementation, the acquisition module 610 is specifically used for: Based on the visual occlusion area and the current distance, the initial reflection area on other target vehicles is determined, wherein the area size of the visual occlusion area is positively correlated with the area size of the initial reflection area, and the current distance is positively correlated with the area size of the initial reflection area; The initial reflection area is adjusted based on the current road slope of other target vehicles to obtain the target reflection area.

[0159] In one possible implementation, the acquisition module 610 is specifically used for: Based on the current road slope, determine the target adjustment amount, where the target adjustment amount is positively correlated with the current road slope; Based on the target adjustment amount, the initial reflection area is adjusted to obtain the target reflection area.

[0160] In one possible implementation, the acquisition module 610 is specifically used for: Based on the target adjustment amount, the initial reflection area is adjusted to obtain the adjusted reflection area; The region of interest in the adjusted reflection region is determined as the target reflection region; Among them, the probability of visual information in the visually occluded area of ​​the region of interest is greater than the probability of visual information in the visually occluded area of ​​the non-region of interest in the adjusted reflection area.

[0161] In one possible implementation, the processing module 620 is specifically used for: Based on the target object's current motion information, determine the target object's current motion intention; Based on the vehicle's current operating information, determine the vehicle's current driving intention; Based on the target object's current motion intention and the vehicle's current driving intention, predict the current collision risk.

[0162] In one possible implementation, the processing module 620 is specifically used for: Based on the geometric dimensions of the visual occlusion area, the target reflection area on other target vehicles is determined. The geometric dimensions of the target reflection area are positively correlated with the geometric dimensions of the visual occlusion area, and there is a corresponding relationship between the geometric dimensions of the visual occlusion area and the target reflection area. The reflected image is determined based on visual information in the target's reflective area.

[0163] In one possible implementation, the processing module 620 is specifically used for: Based on the target object's current motion information, determine the target object's current motion intention; Predict the current collision risk based on the target object's current movement intention.

[0164] In one possible implementation, the processing module 620 is specifically used for: Based on the vehicle's current operating information, determine the vehicle's current driving intention; Based on the target object's current motion intention and the vehicle's current driving intention, predict the current collision risk.

[0165] In one possible implementation, the processing module 620 is specifically used for: When there is only one target object, the current collision risk is predicted based on the current motion intention of the target object and the current driving intention of the vehicle. When there are multiple target objects, a first target object is determined based on the current distance between each target object and the vehicle; the current collision risk is predicted based on the current movement intention of the first target object and the current driving intention of the vehicle, where the current distance between the first target object and the vehicle is the smallest; or, When there are multiple target objects, the collision risk between the vehicle and each target object is determined based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle. The collision risks of multiple targets are fused to obtain the current collision risk.

[0166] In one possible implementation, the processing module 620 is specifically used for: Based on the current motion intention of each target object and the current driving intention of the vehicle, the initial collision risk between the vehicle and each target object is determined; Based on the current distance between each target object and the vehicle, the weight of each target object is determined, wherein the weight of each target object is negatively correlated with the current distance between each target object and the vehicle; The initial collision risk between the vehicle and each target object is weighted by the weight of each target object to obtain the target collision risk between the vehicle and each target object.

[0167] It should be noted that the aforementioned device 600 is embodied in the form of a functional module. The term "module" here can be implemented in software and / or hardware, without specific limitations.

[0168] For example, a "module" can be a software program, hardware circuit, or a combination of both that implements the above functions. Hardware circuits may include application-specific integrated circuits (ASICs), electronic circuits, processors (e.g., shared processors, proprietary processors, or combined processors) and memory for executing one or more software or firmware programs, combined logic circuits, and / or other suitable components that support the described functions.

[0169] Therefore, the modules of the various examples described in the embodiments of this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0170] Figure 7 This is a schematic diagram of the controller provided in the embodiments of this application.

[0171] For example, such as Figure 7 As shown, the vehicle includes a controller 700, which includes a storage module 710 and a processing module 720. The storage module 710 stores executable program code 7101, and the processing module 720 is used to call and execute the executable program code 7101 to perform a vehicle collision avoidance warning method.

[0172] Figure 8 This is a schematic diagram of the vehicle structure provided in the embodiments of this application.

[0173] For example, such as Figure 8 As shown, the vehicle 800 includes a memory 810 and a processor 820. The memory 810 stores executable program code 8101, and the processor 820 is used to call and execute the executable program code 8101 to perform a vehicle collision avoidance warning method.

[0174] This application can divide the vehicle into functional modules based on the above method example. For example, each module can correspond to a separate function module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0175] When each functional module is divided according to its corresponding function, the vehicle may include: an acquisition module and a processing module, etc. It should be noted that all relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0176] The vehicle provided in this application is used to execute the aforementioned vehicle collision avoidance warning method, and thus can achieve the same effect as the aforementioned implementation method.

[0177] When using integrated units, the vehicle may include a processing module and a storage module. The processing module is used to control and manage the vehicle's movements. The storage module is used to support the vehicle in executing relevant program code and data.

[0178] The processing module may be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits shown in conjunction with the disclosure of this application. The processor may also be a combination of functions that implement computing capabilities, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and microprocessors, etc., and the storage module may be a memory.

[0179] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described in the foregoing embodiments. The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs (Digital Video Discs), CD-ROMs (Compact Disc Read-Only Memory), microdrives, magneto-optical disks, ROMs (Read-Only Memory), RAMs (Random Access Memory), EPROMs (Erasable Programmable Read-Only Memory), EEPROMs (Electrically Erasable Programmable Read Only Memory), DRAMs (Dynamic Random Access Memory), VRAMs (Video Random Access Memory), flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.

[0180] This application also provides a computer program product that, when run on a computer, causes the computer to perform the aforementioned steps to implement a vehicle collision avoidance warning method as described in the above embodiments.

[0181] In addition, the vehicle provided in the embodiments of this application may specifically be a chip, component or module. The vehicle may include a connected processor and a memory. The memory is used to store instructions. When the vehicle is running, the processor can call and execute the instructions to make the chip execute a vehicle collision avoidance warning method in the above embodiments.

[0182] The vehicle, computer-readable storage medium, computer program product or chip provided in this application are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0183] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0184] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0185] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A vehicle collision avoidance warning method, characterized in that, The method includes: When a visual obstruction area is detected in front of the vehicle, a reflected image on another target vehicle is acquired based on the visual obstruction area. The reflected image is used to represent the visual image reflected from the visual obstruction area to the other target vehicle. The visual obstruction area is caused by an external obstruction of the vehicle. In response to the presence of a target object in the reflected image, based on the current motion information of the target object, the current collision risk between the vehicle and the target object is predicted, wherein the target object is used to represent an object that obstructs the vehicle's movement in the visual occlusion area; Based on the current collision risk, the vehicle is controlled to issue a collision avoidance warning.

2. The method according to claim 1, characterized in that, The step of acquiring reflected images of other target vehicles based on the visual occlusion area includes: Obtain the current distance between the other target vehicles and the vehicle; Based on the visual occlusion area and the current distance, the target reflection area on the other target vehicle is determined, wherein the area size of the target reflection area is positively correlated with the area size of the visual occlusion area, and the area size of the target reflection area is positively correlated with the current distance; The reflected image is determined based on visual information in the target reflection area.

3. The method according to claim 2, characterized in that, Determining the target reflection area on the other target vehicles based on the visual occlusion area and the current distance includes: Based on the visual occlusion area and the current distance, an initial reflection area on the other target vehicle is determined, wherein the area size of the initial reflection area is positively correlated with the area size of the visual occlusion area, and the area size of the initial reflection area is positively correlated with the current distance; Based on the current road slope of the other target vehicles, the initial reflection area is adjusted to obtain the target reflection area.

4. The method according to claim 3, characterized in that, The process of adjusting the size of the initial reflection area based on the current road slope of the other target vehicles to obtain the target reflection area includes: Based on the current road slope, a target adjustment amount is determined, wherein the target adjustment amount is positively correlated with the current road slope; Based on the target adjustment amount, the initial reflection area is adjusted to obtain the target reflection area.

5. The method according to claim 4, characterized in that, The step of adjusting the initial reflection region based on the target adjustment amount to obtain the target reflection region includes: Based on the target adjustment amount, the initial reflection area is adjusted to obtain the adjusted reflection area; The region of interest in the adjusted reflection region is determined as the target reflection region; The probability of visual information in the visually occluded region within the region of interest is greater than the probability of visual information in the visually occluded region within the non-interested region of the adjusted reflection region.

6. The method according to claim 1, characterized in that, The step of acquiring reflected images of other target vehicles based on the visual occlusion area includes: Based on the geometric dimensions of the visual occlusion area, the target reflection area on the other target vehicle is determined, wherein the geometric dimensions of the target reflection area are positively correlated with the geometric dimensions of the visual occlusion area, and there is a corresponding relationship between the geometric dimensions of the visual occlusion area and the target reflection area; The reflected image is determined based on visual information in the target reflection area.

7. The method according to any one of claims 1 to 6, characterized in that, The step of predicting the current collision risk between the vehicle and the target object based on the target object's current motion information includes: Based on the current motion information of the target object, determine the current motion intention of the target object; Based on the vehicle's current operating information, determine the vehicle's current driving intention; The current collision risk is predicted based on the current movement intention of the target object and the current driving intention of the vehicle.

8. The method according to claim 7, characterized in that, The prediction of the current collision risk based on the current motion intention of the target object and the current driving intention of the vehicle includes: When there is only one target object, the current collision risk is predicted based on the current movement intention of the target object and the current driving intention of the vehicle. When there are multiple target objects, a first target object is determined based on the current distance between each target object and the vehicle; the current collision risk is predicted based on the current movement intention of the first target object and the current driving intention of the vehicle, wherein the current distance between the first target object and the vehicle is the smallest; or... When there are multiple target objects, the target collision risk between the vehicle and each target object is determined based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle. The current collision risk is obtained by performing risk fusion on multiple target collision risks.

9. The method according to claim 8, characterized in that, The determination of the target collision risk between the vehicle and each of the target objects based on the current distance between each target object and the vehicle, the current movement intention of each target object, and the current driving intention of the vehicle includes: Based on the current motion intention of each of the target objects and the current driving intention of the vehicle, the initial collision risk between the vehicle and each of the target objects is determined; Based on the current distance between each target object and the vehicle, a weight is determined for each target object, wherein the weight of each target object is negatively correlated with the current distance between each target object and the vehicle; The initial collision risk between the vehicle and each of the target objects is weighted and the weight of each of the target objects is calculated to obtain the target collision risk between the vehicle and each of the target objects.

10. A vehicle, characterized in that, The vehicles include: Memory, used to store executable program code; A processor for calling and running the executable program code from the memory, causing the vehicle to perform the method as described in any one of claims 1 to 9.

Images