Parking distance control method, storage medium and device including visual collision monitoring

Abstract

The present invention discloses a parking distance control method, storage medium and device including visual collision monitoring, which relates to the field of intelligent vehicle control. The method includes obtaining the distance information between the vehicle and the obstacle based on the ultrasonic radar arranged around the vehicle, and obtaining the PDC recognition result; obtaining the obstacle information based on the collected vehicle surrounding environment image recognition, and performing collision detection according to the obstacle information and the vehicle running trajectory; weighting and fusing the PDC recognition result and the collision detection result to obtain the final result for collision prompting. The present invention combines the obstacle distance information of the ultrasonic radar, and the two results are weighted and fused to obtain the final result and perform collision prompting, effectively avoiding the problem of missed recognition caused by unilateral failure of the radar or surround recognition.

Description

Technical Field

[0001] The present invention relates to the field of vehicle intelligent control, and in particular to a parking distance control method, storage medium and device including visual collision monitoring. Background Art

[0002] The automatic parking assist system uses a fusion of vision and ultrasonic radar to identify parking spaces and obstacle information, thereby completing the automatic parking of the vehicle through the control and actuator. Specifically, the 4-channel camera inputs the image to the APA (Auto Parking Assist System) controller, and the image stitching algorithm completes the surround image stitching. The perception algorithm completes the identification of parking spaces, obstacles, wheel blocks, ground locks and other targets on the stitched image; the ultrasonic radar can complete the relative distance between obstacles and vehicles and spatial parking space identification through ranging. Finally, the fusion module completes the fusion of parking space information and outputs the obstacle information to the control module, which completes the path planning and lateral and longitudinal control, and finally controls the vehicle actuator to complete the entire parking action.

[0003] At the same time, the ultrasonic radar also provides guidance for the driver's parking and exit operations when the vehicle is not performing automatic parking. This function is called PDC (Parking Distance Control System), but the PDC function on the market often only involves ultrasonic radar when working.

[0004] With the promotion of PDC, drivers will rely more and more on PDC when driving at low speeds or passing through areas close to obstacles. However, when the ultrasonic wave fails or is in the blind spot, PDC cannot provide the driver with better guidance to stay away from obstacles, and it is very easy for the vehicle to scratch against obstacles. Summary of the invention

[0005] In view of the defects existing in the prior art, the purpose of the present invention is to provide a parking distance control method, storage medium and device including visual collision monitoring, which combines the obstacle distance information of the ultrasonic radar, and weightedly fuses the two results to obtain the final result and provide a collision prompt, thereby effectively avoiding the problem of missed recognition caused by unilateral failure of radar or surround view recognition.

[0006] To achieve the above objectives, the present invention provides a parking distance control method including visual collision monitoring, which specifically includes the following steps:

[0007] The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, and obtains the PDC recognition result;

[0008] Obstacle information is obtained based on the collected image recognition of the vehicle's surrounding environment, and collision detection is performed based on the obstacle information and the vehicle's running trajectory;

[0009] The PDC recognition results and collision detection results are weighted and fused to obtain the final result for collision prompting.

[0010] On the basis of the above technical solution, the ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle and obtains the PDC recognition result. The specific steps include:

[0011] Based on the coverage of ultrasonic radars arranged around the vehicle, the area around the vehicle is divided into multiple PDC areas;

[0012] According to the detection of the ultrasonic radar, the distance information between the obstacle and the vehicle in each PDC area is obtained to obtain the PDC recognition result.

[0013] On the basis of the above technical solution, the obstacle information is obtained based on the collected vehicle surrounding environment image recognition, and the specific steps include:

[0014] The camera collects the surround view image or the front view image of the vehicle to form the environment image around the vehicle;

[0015] The deep learning network model is used to extract semantic information from the surround view image and identify obstacles around the vehicle.

[0016] On the basis of the above technical solution, the collision detection is performed according to the obstacle information and the vehicle running trajectory, wherein the collision detection method includes a projection method and a triangular area method.

[0017] On the basis of the above technical solution, the collision detection between obstacles and vehicle running tracks is carried out by using the triangular area method. The specific steps include:

[0018] Calculate the coordinates of each vertex of the body geometry:

[0019] X A =xl r *cosθ+(l w / 2)*sinθ

[0020] Y A =yl r *sinθ-(l w / 2)*cosθ

[0021] X B =x+l f *cosθ+(l w / 2)*sinθ

[0022] Y B =x+l f *sinθ-(l w / 2)*cosθ

[0023] X C =x+l f *cosθ-(l w / 2)*sinθ

[0024] Y C =x+l f *sinθ+(l w / 2)*cosθ

[0025] X D =xl r *cosθ-(l w / 2)*sinθ

[0026] Y D =xl r *cosθ+(l w / 2)*cosθ

[0027] Among them, X A Indicates the horizontal coordinate of the vertex of the lower right corner of the body geometry, Y A Indicates the vertical coordinate of the vertex of the lower right corner of the body geometry, X B Indicates the horizontal coordinate of the vertex of the upper right corner of the body geometry, Y B Indicates the vertical coordinate of the vertex of the upper right corner of the body geometry contour, X C Indicates the horizontal coordinate of the upper left corner of the body geometry, Y C Indicates the ordinate of the upper left corner of the body geometry, X D Indicates the horizontal coordinate of the vertex of the lower left corner of the body geometry, Y D represents the ordinate of the vertex at the lower left corner of the vehicle body geometric outline, x represents the abscissa of the path planning point, y represents the ordinate of the path planning point, l w Indicates vehicle width, l r Represents the rear overhang length, l f represents the sum of the rear overhang length and the wheelbase, θ represents the heading angle;

[0028] According to the corner points in the parking space and the vertices of the geometric contour of the vehicle body, the corner point area is calculated. The specific calculation method is:

[0029] S1=abs(((X A -X Point )*(Y A -Y Point )-(Y B -Y Point )*(XB -X Point )) / 2)

[0030] S2=abs(((X B -X Point )*(Y C -Y Point )-(Y B -Y Point )*(X C -X Point )) / 2)

[0031] S3=abs(((X C -X Point )*(Y D -Y Point )-(Y C -Y Point )*(X D -X Point )) / 2)

[0032] S4=abs(((X D -X Point )*(Y A -Y Point )-(Y D -Y Point )*(X A -X Point )) / 2)

[0033] Among them, S1 represents the corner point area of ​​the triangle formed by the lower right corner vertex of the body geometric outline, the upper right corner vertex of the body geometric outline and the corner point inside the parking space, S2 represents the corner point area of ​​the triangle formed by the upper right corner vertex of the body geometric outline, the upper left corner vertex of the body geometric outline and the corner point inside the parking space, S3 represents the corner point area of ​​the triangle formed by the upper left corner vertex of the body geometric outline, the lower left corner vertex of the body geometric outline and the corner point inside the parking space, S4 represents the corner point area of ​​the triangle formed by the lower left corner vertex of the body geometric outline, the lower right corner vertex of the body geometric outline and the corner point inside the parking space, X Point Indicates the horizontal coordinate of the corner point inside the parking space, Y Point Indicates the ordinate of the corner point in the parking space, abs indicates absolute value calculation;

[0034] Based on the comparison between the sum of the areas of all corner points and the geometric contour area of ​​the vehicle body, a judgment is made as to whether a collision occurs during path planning, and the collision detection result is obtained.

[0035] On the basis of the above technical solution, the PDC recognition result and the collision detection result are weighted and merged to obtain the final result for collision prompting. The specific steps include:

[0036] Based on the blind spot range of the vehicle, set the weights of the PDC recognition results and collision detection results in different blind spots;

[0037] According to the set weights, the PDC recognition results and collision detection results are weighted and fused to obtain the final result for collision prompting.

[0038] On the basis of the above technical solution, the final result is obtained to provide a collision prompt, specifically:

[0039] The final result is sent to the vehicle's IVI or sound system for collision prompting.

[0040] On the basis of the above technical solution, before the ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, it also includes:

[0041] The vehicle speed is obtained, and when the vehicle speed meets the set low-speed range and the vehicle PDC activation conditions are met, the distance information between the vehicle and the obstacle is obtained based on the ultrasonic radar arranged around the vehicle.

[0042] The present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned parking distance control method including visual collision monitoring are implemented.

[0043] The present invention provides a parking distance control device including visual collision monitoring, comprising:

[0044] Ultrasonic sensing module, which is used to obtain the distance information between the vehicle and the obstacle based on the ultrasonic radar arranged around the vehicle, and obtain the PDC recognition result;

[0045] A visual perception module, which is used to obtain obstacle information based on the collected images of the vehicle's surrounding environment, and perform collision detection based on the obstacle information and the vehicle's running trajectory;

[0046] The fusion judgment module is used to weight and fuse the PDC recognition result and the collision detection result to obtain the final result for collision prompting.

[0047] Compared with the prior art, the advantages of the present invention are: providing a strategy for visual and ultrasonic radar fusion monitoring, that is, when PDC starts working, the surround-view camera will also process the obstacle information around the vehicle in real time, mark the surrounding obstacle information through target recognition and semantic segmentation, and combine the obstacle distance information of the ultrasonic radar, and weightedly fuse the two results to obtain the final result and provide a collision prompt, effectively avoiding the problem of missed recognition caused by unilateral failure of radar or surround-view recognition, and avoiding the risk of collision caused by the close distance between the vehicle and the obstacle due to blind spots or sensor failures. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 A flowchart of a parking distance control method including visual collision monitoring in an embodiment of the present invention;

[0050] Figure 2 It is a schematic diagram of the division of the PDC area;

[0051] Figure 3 It is a schematic diagram of collision detection using the triangular area method;

[0052] Figure 4 The figure is a schematic diagram of the structure of a parking distance control device including visual collision monitoring in the present invention. DETAILED DESCRIPTION

[0053] The embodiment of the present invention provides a parking distance control method including visual collision monitoring, and provides a strategy for visual and ultrasonic radar fusion monitoring, that is, when PDC starts working, the surround view camera will also process the obstacle information around the vehicle in real time, mark the surrounding obstacle information in the form of target recognition and semantic segmentation, and combine the obstacle distance information of the ultrasonic radar. The two results are weighted and fused to obtain the final result and provide a collision prompt, effectively avoiding the problem of missed recognition caused by unilateral failure of radar or surround view recognition, and avoiding the risk of collision caused by the close distance between the vehicle and the obstacle due to blind spots or sensor failures. The embodiment of the present invention also provides a non-transient computer-readable storage medium and a parking distance control device including visual collision monitoring.

[0054] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0055] See also Figure 1 As shown, a parking distance control method including visual collision monitoring provided by an embodiment of the present invention specifically includes the following steps:

[0056] S1: The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, and obtains the PDC recognition result; the PDC recognition result is used to indicate the possibility of collision between the vehicle and the obstacle when traveling on the planned path.

[0057] In the present invention, the distance information between the vehicle and the obstacle is obtained based on the ultrasonic radar arranged around the vehicle, and the PDC recognition result is obtained. The specific steps include:

[0058] S101: Divide the area around the vehicle into a plurality of PDC areas based on the coverage of ultrasonic radars arranged around the vehicle;

[0059] S102: According to the detection of the ultrasonic radar, the distance information between the obstacle and the vehicle in each PDC area is obtained to obtain a PDC recognition result.

[0060] That is, the area around the vehicle is divided into multiple PDC areas. For each PDC area, based on the detection results of the ultrasonic radar, the possibility of collision between the vehicle and the obstacle on the planned path is judged to obtain the PDC recognition result of the current area.

[0061] See also Figure 2 As shown, in a possible implementation, 12 ultrasonic radars are arranged around the vehicle, and the area around the vehicle body is divided into 16 PDC areas according to the actual installation area or virtual area of ​​the ultrasonic radars. Figure 2 In the figure, numbers 1 to 16 represent 16 PDC areas, among which 13, 14, 15, and 16 are virtual SDW (safety distance warning) sensing areas, and distance information is obtained from the ultrasonic sensor drive according to the PDC area.

[0062] Furthermore, the ultrasonic radar ranging information of the 16 PDC areas may be marked in a certain corresponding manner.

[0063] S2: Obstacle information is obtained based on the collected image of the vehicle's surrounding environment, and collision detection is performed based on the obstacle information and the vehicle's running trajectory;

[0064] In the present invention, obstacle information is obtained based on the collected vehicle surrounding environment image recognition, and the specific steps include:

[0065] S201: acquiring a surround view image or a front view image of the vehicle through a camera to form an environment image around the vehicle;

[0066] S202: Extracting semantic information from the surround view image through a deep learning network model to identify obstacles around the vehicle.

[0067] That is, the camera installed on the vehicle collects images around the vehicle to obtain surround view images or front view images, thereby obtaining the environment image around the vehicle, and then extracts semantic information from the surround view image through deep learning network models such as Unet and DeepLab to identify obstacles in the image. Specifically, obstacles are other vehicles, pillars, pedestrians, non-motor vehicle walls, etc.

[0068] After the obstacle information is identified, collision detection can be performed based on the obstacle information and the vehicle running track. In the present invention, the collision detection method includes a projection method and a triangular area method.

[0069] See also Figure 3 As shown, the collision detection between obstacles and vehicle running tracks using the triangular area method is specifically described below.

[0070] a: Calculate the coordinates of each vertex of the body geometry:

[0071] X A =xl r *cosθ+(l w / 2)*sinθ

[0072] Y A =yl r *sinθ-(l w / 2)*cosθ

[0073] X B =x+l f *cosθ+(l w / 2)*sinθ

[0074] Y B =x+l f *sinθ-(l w / 2)*cosθ

[0075] X C =x+l f *cosθ-(l w / 2)*sinθ

[0076] Y C =x+l f *sinθ+(l w / 2)*cosθ

[0077] X D =xl r *cosθ-(l w / 2)*sinθ

[0078] Y D =xl r*cosθ+(l w / 2)*cosθ

[0079] Among them, X A Indicates the horizontal coordinate of the vertex of the lower right corner of the body geometry, Y A Indicates the vertical coordinate of the vertex of the lower right corner of the body geometry, X B Indicates the horizontal coordinate of the vertex of the upper right corner of the body geometry, Y B Indicates the vertical coordinate of the vertex of the upper right corner of the body geometry contour, X C Indicates the horizontal coordinate of the upper left corner of the body geometry, Y C Indicates the ordinate of the upper left corner of the body geometry, X D Indicates the horizontal coordinate of the vertex of the lower left corner of the body geometry, Y D represents the ordinate of the vertex at the lower left corner of the vehicle body geometric outline, x represents the abscissa of the path planning point, y represents the ordinate of the path planning point, l w Indicates vehicle width, l r Represents the rear overhang length, l f represents the sum of the rear overhang length and the wheelbase, θ represents the heading angle; Figure 3 In the figure, point A is the vertex of the lower right corner of the body geometric outline, point B is the vertex of the upper right corner of the body geometric outline, point C is the vertex of the upper left corner of the body geometric outline, point D is the vertex of the lower left corner of the body geometric outline, and point Point is the inner corner point of the parking space.

[0080] b: Calculate the corner area based on the corner points in the parking space and the vertices of the geometric contour of the vehicle body. The specific calculation method is:

[0081] S1=abs(((X A -X Point )*(Y A -Y Point )-(Y B -Y Point )*(X B -X Point )) / 2)

[0082] S2=abs(((X B -X Point )*(Y C -Y Point )-(Y B -Y Point )*(X C -X Point )) / 2)

[0083] S3=abs(((X C -X Point )*(Y D -Y Point )-(YC -Y Point )*(X D -X Point )) / 2)

[0084] S4=abs(((X D -X Point )*(Y A -Y Point )-(Y D -Y Point )*(X A -X Point )) / 2)

[0085] Among them, S1 represents the corner point area of ​​the triangle formed by the lower right corner vertex of the body geometric outline, the upper right corner vertex of the body geometric outline and the corner point inside the parking space, S2 represents the corner point area of ​​the triangle formed by the upper right corner vertex of the body geometric outline, the upper left corner vertex of the body geometric outline and the corner point inside the parking space, S3 represents the corner point area of ​​the triangle formed by the upper left corner vertex of the body geometric outline, the lower left corner vertex of the body geometric outline and the corner point inside the parking space, S4 represents the corner point area of ​​the triangle formed by the lower left corner vertex of the body geometric outline, the lower right corner vertex of the body geometric outline and the corner point inside the parking space, X Point Indicates the horizontal coordinate of the corner point inside the parking space, Y Point Indicates the ordinate of the corner point in the parking space, abs indicates absolute value calculation;

[0086] c: Based on the comparison between the sum of the areas of all corner points and the area of ​​the geometric contour of the vehicle body, a judgment is made as to whether a collision occurs during path planning, and the collision detection result is obtained.

[0087] That is, first calculate the sum of S1 to S4 to obtain the areas of all corner points, then calculate the difference between the areas of all corner points and the area of ​​the vehicle body geometric contour, and compare the calculated difference with the set difference range to determine whether there is a possibility of collision between the vehicle and the planned path.

[0088] S3: Weight and fuse the PDC recognition result and the collision detection result to obtain a final result for collision prompting.

[0089] In the present invention, the PDC recognition result and the collision detection result are weighted and fused to obtain the final result for collision prompting. The specific steps include:

[0090] S301: Based on the blind spot range of the vehicle, weights of the PDC recognition results and collision detection results in different blind spots are set; for example, for the blind spot directly behind the vehicle, the weight coefficient of the PDC recognition result is set to 0.5, and the weight coefficient of the collision detection result is set to 0.5;

[0091] S302: According to the set weight, the PDC recognition result and the collision detection result are weighted and fused to obtain a final result for collision prompting, that is, the PDC recognition result and the collision detection result are multiplied by the corresponding weight coefficients respectively and then added to obtain the final result.

[0092] It should be noted that the PDC recognition result is the possibility of a collision. The larger the value, the greater the possibility of a collision. The collision detection result is also the possibility of a collision. The larger the value, the greater the possibility of a collision. For the final result obtained, the larger the value, the greater the possibility of a collision.

[0093] In the present invention, the final result is obtained to provide a collision warning, specifically: the final result is sent to the vehicle's IVI (in-vehicle infotainment) or sound system to provide a collision warning. That is, the final result is sent to the IVI or sound system to provide a collision risk warning to the driver.

[0094] In the present invention, before the ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, it also includes: obtaining the speed of the vehicle, and when the speed of the vehicle meets the set low-speed range and the vehicle PDC starting conditions are met (that is, the current driving scene meets the PDC starting conditions, such as reversing, etc.), the ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle.

[0095] In the parking distance control method including visual collision monitoring of the present invention, PDC performs visual target detection and semantic segmentation while measuring distance, realizes visual obstacle detection, and the detection results are fused in blind areas according to the recognition confidence. During the low-speed driving of the vehicle, the ultrasonic radar ranging information and the visual obstacle information are fused in real time according to the set weights to obtain the final detection result.

[0096] For the PDC alarm function, the accuracy of ultrasonic ranging is greatly affected by the ambient temperature. Generally, the original radar distance information is obtained from the vehicle and then output to the ultrasonic sensing module after temperature compensation. The sensing algorithm verifies the sensing performance by means of threshold filtering, echo width filtering, echo assignment filtering, secondary comparison filtering, tracking filtering, etc. on the ultrasonic radar data. The verified information then obtains the height information of the obstacle according to the corresponding relationship between the number of echoes and the distance, and then the tracking algorithm obtains slightly more accurate obstacle position and size information. In the present invention, after the vehicle meets the operating conditions during low-speed driving, the surround semantic target recognition and semantic segmentation begin to run, and the ultrasonic perception and visual perception are weighted according to the corresponding confidence values. The visual obstacle determines whether there is a collision risk by the overlapping area with the geometric model of the vehicle; the ultrasonic perception determines the collision risk by the distance from the obstacle. The two assist each other to inform the user of the warning information through IVI to avoid the collision risk caused by the close distance to the obstacle due to blind spots or sensor failures.

[0097] The parking distance control method including visual collision monitoring in the embodiment of the present invention provides a strategy of visual and ultrasonic radar fusion monitoring, that is, when PDC starts working, the surround-view camera will also process the obstacle information around the vehicle in real time, and mark the surrounding obstacle information in the form of target recognition and semantic segmentation. Combined with the obstacle distance information of the ultrasonic radar, the two results are weightedly fused to obtain the final result and a collision prompt is given, effectively avoiding the problem of missed recognition caused by unilateral failure of radar or surround-view recognition, and avoiding the risk of collision caused by the close distance between the vehicle and the obstacle due to blind spots or sensor failures.

[0098] In a possible implementation, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, the readable storage medium is located in a PLC (Programmable Logic Controller) controller, and a computer program is stored on the readable storage medium. When the program is executed by a processor, the following steps of the parking distance control method including visual collision monitoring are implemented:

[0099] The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, and obtains the PDC recognition result;

[0100] Obstacle information is obtained based on the collected image recognition of the vehicle's surrounding environment, and collision detection is performed based on the obstacle information and the vehicle's running trajectory;

[0101] The PDC recognition results and collision detection results are weighted and fused to obtain the final result for collision prompting.

[0102] The storage medium may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, device or device.

[0103] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0104] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0105] See also Figure 4As shown, an embodiment of the present invention provides a parking distance control device including visual collision monitoring, including an ultrasonic sensing module, a visual sensing module and a fusion judgment module.

[0106] The ultrasonic sensing module is used to obtain the distance information between the vehicle and the obstacle based on the ultrasonic radar arranged around the vehicle, and obtain the PDC recognition result; the visual perception module is used to obtain the obstacle information based on the collected image recognition of the vehicle's surrounding environment, and perform collision detection based on the obstacle information and the vehicle's running trajectory; the fusion judgment module is used to weight and fuse the PDC recognition result and the collision detection result to obtain the final result for collision prompt.

[0107] In the present invention, the distance information between the vehicle and the obstacle is obtained based on the ultrasonic radar arranged around the vehicle, and the PDC recognition result is obtained. The specific steps include:

[0108] Based on the coverage of ultrasonic radars arranged around the vehicle, the area around the vehicle is divided into multiple PDC areas;

[0109] According to the detection of the ultrasonic radar, the distance information between the obstacle and the vehicle in each PDC area is obtained to obtain the PDC recognition result.

[0110] In the present invention, obstacle information is obtained based on the collected vehicle surrounding environment image recognition, and the specific steps include:

[0111] The camera collects the surround view image or the front view image of the vehicle to form the environment image around the vehicle;

[0112] The deep learning network model is used to extract semantic information from the surround view image and identify obstacles around the vehicle.

[0113] In the present invention, collision detection is performed based on obstacle information and vehicle running trajectory, wherein the collision detection method includes a projection method and a triangular area method.

[0114] In the present invention, the triangular area method is used to detect the collision between the obstacle and the vehicle running track, and the specific steps include:

[0115] Calculate the coordinates of each vertex of the body geometry:

[0116] X A =xl r *cosθ+(l w / 2)*sinθ

[0117] Y A =yl r *sinθ-(l w / 2)*cosθ

[0118] X B =x+l f *cosθ+(l w / 2)*sinθ

[0119] Y B =x+l f *sinθ-(l w / 2)*cosθ

[0120] X C =x+l f *cosθ-(l w / 2)*sinθ

[0121] Y C =x+l f *sinθ+(l w / 2)*cosθ

[0122] X D =xl r *cosθ-(l w / 2)*sinθ

[0123] Y D =xl r *cosθ+(l w / 2)*cosθ

[0124] Among them, X A Indicates the horizontal coordinate of the vertex of the lower right corner of the body geometry, Y A Indicates the vertical coordinate of the vertex of the lower right corner of the body geometry, X B Indicates the horizontal coordinate of the vertex of the upper right corner of the body geometry, Y B Indicates the vertical coordinate of the vertex of the upper right corner of the body geometry contour, X C Indicates the horizontal coordinate of the upper left corner of the body geometry, Y C Indicates the ordinate of the upper left corner of the body geometry, X D Indicates the horizontal coordinate of the vertex of the lower left corner of the body geometry, Y D represents the ordinate of the vertex at the lower left corner of the vehicle body geometric outline, x represents the abscissa of the path planning point, y represents the ordinate of the path planning point, l w Indicates vehicle width, l r Represents the rear overhang length, l f represents the sum of the rear overhang length and the wheelbase, θ represents the heading angle;

[0125] According to the corner points in the parking space and the vertices of the geometric contour of the vehicle body, the corner point area is calculated. The specific calculation method is:

[0126] S1=abs(((X A -XPoint )*(Y A -Y Point )-(Y B -Y Point )*(X B -X Point )) / 2)

[0127] S2=abs(((X B -X Point )*(Y C -Y Point )-(Y B -Y Point )*(X C -X Point )) / 2)

[0128] S3=abs(((X C -X Point )*(Y D -Y Point )-(Y C -Y Point )*(X D -X Point )) / 2)

[0129] S4=abs(((X D -X Point )*(Y A -Y Point )-(Y D -Y Point )*(X A -X Point )) / 2)

[0130] Among them, S1 represents the corner point area of ​​the triangle formed by the lower right corner vertex of the body geometric outline, the upper right corner vertex of the body geometric outline and the corner point inside the parking space, S2 represents the corner point area of ​​the triangle formed by the upper right corner vertex of the body geometric outline, the upper left corner vertex of the body geometric outline and the corner point inside the parking space, S3 represents the corner point area of ​​the triangle formed by the upper left corner vertex of the body geometric outline, the lower left corner vertex of the body geometric outline and the corner point inside the parking space, S4 represents the corner point area of ​​the triangle formed by the lower left corner vertex of the body geometric outline, the lower right corner vertex of the body geometric outline and the corner point inside the parking space, X Point Indicates the horizontal coordinate of the corner point inside the parking space, Y Point Indicates the ordinate of the corner point in the parking space, abs indicates absolute value calculation;

[0131] Based on the comparison between the sum of the areas of all corner points and the geometric contour area of ​​the vehicle body, a judgment is made as to whether a collision occurs during path planning, and the collision detection result is obtained.

[0132] In the present invention, the PDC recognition result and the collision detection result are weighted and fused to obtain the final result for collision prompting. The specific steps include:

[0133] Based on the blind spot range of the vehicle, set the weights of the PDC recognition results and collision detection results in different blind spots;

[0134] According to the set weights, the PDC recognition results and collision detection results are weighted and fused to obtain the final result for collision prompting.

[0135] In the present invention, the final result is obtained to provide a collision prompt, specifically:

[0136] The final result is sent to the vehicle's IVI or sound system for collision prompting.

[0137] In the present invention, before the ultrasonic radar arranged around the vehicle acquires the distance information between the vehicle and the obstacle, the following steps are also included:

[0138] The vehicle speed is obtained, and when the vehicle speed meets the set low-speed range and the vehicle PDC activation conditions are met, the distance information between the vehicle and the obstacle is obtained based on the ultrasonic radar arranged around the vehicle.

[0139] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

[0140] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the process in the flowchart. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

Claims

1. A parking distance control method including visual collision monitoring, characterized in that: The specific steps include: The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, and obtains the PDC recognition result; Obstacle information is obtained based on the collected image recognition of the vehicle's surrounding environment, and collision detection is performed based on the obstacle information and the vehicle's running trajectory; The PDC recognition results and collision detection results are weighted and merged to obtain the final result for collision prompting; The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle and obtains the PDC recognition result. The specific steps include: Based on the coverage of ultrasonic radars arranged around the vehicle, the area around the vehicle is divided into multiple PDC areas; According to the detection of ultrasonic radar, the distance information between the obstacle and the vehicle in each PDC area is obtained to obtain the PDC recognition result; Wherein, the collision detection is performed based on the obstacle information and the vehicle running trajectory, wherein the collision detection method includes a projection method and a triangular area method.

2. A parking distance control method including visual collision monitoring as claimed in claim 1, characterized in that: Obtaining obstacle information based on the collected vehicle surrounding environment image recognition specifically includes: The camera collects the surround view image or the front view image of the vehicle to form the environment image around the vehicle; The deep learning network model is used to extract semantic information from the surround view image and identify obstacles around the vehicle.

3. A parking distance control method including visual collision monitoring as claimed in claim 1, characterized in that: The triangular area method is used to detect the collision between obstacles and vehicle running tracks. The specific steps include: Calculate the coordinates of each vertex of the body geometry: in, Represents the horizontal coordinate of the vertex of the lower right corner of the body geometry contour, Indicates the vertical coordinate of the vertex of the lower right corner of the body geometry contour, Represents the horizontal coordinate of the vertex of the upper right corner of the body geometry contour, Indicates the ordinate of the vertex of the upper right corner of the body geometry contour, Represents the horizontal coordinate of the vertex of the upper left corner of the body geometry contour, Indicates the ordinate of the upper left corner vertex of the body geometry contour, Represents the horizontal coordinate of the vertex of the lower left corner of the body geometry contour, Indicates the ordinate of the vertex of the lower left corner of the body geometry contour, Represents the horizontal coordinate of the path planning point, Represents the ordinate of the path planning point, Indicates the width of the vehicle. Represents the length of the rear overhang, It represents the sum of rear overhang length and wheelbase. Indicates the heading angle; According to the corner points in the parking space and the vertices of the geometric contour of the vehicle body, the corner point area is calculated. The specific calculation method is: in, It represents the corner point area of ​​the triangle formed by the lower right corner vertex of the body geometric outline, the upper right corner vertex of the body geometric outline and the corner point inside the parking space. It represents the corner point area of ​​the triangle formed by the upper right corner vertex of the body geometric outline, the upper left corner vertex of the body geometric outline and the corner point inside the parking space. It represents the corner point area of ​​the triangle formed by the upper left corner vertex of the body geometric outline, the lower left corner vertex of the body geometric outline and the corner point inside the parking space. It represents the corner point area of ​​the triangle formed by the lower left corner vertex of the body geometric outline, the lower right corner vertex of the body geometric outline and the corner point inside the parking space. Indicates the horizontal coordinate of the corner point inside the parking space, Indicates the ordinate of the corner point inside the parking space, Indicates absolute value calculation; Based on the comparison between the sum of the areas of all corner points and the geometric contour area of ​​the vehicle body, a judgment is made as to whether a collision occurs during path planning, and the collision detection result is obtained.

4. A parking distance control method including visual collision monitoring as claimed in claim 1, characterized in that: The PDC recognition result and the collision detection result are weighted and merged to obtain the final result for collision prompting, and the specific steps include: Based on the blind spot range of the vehicle, set the weights of the PDC recognition results and collision detection results in different blind spots; According to the set weights, the PDC recognition results and collision detection results are weighted and fused to obtain the final result for collision prompting.

5. A parking distance control method including visual collision monitoring as claimed in claim 1, characterized in that: The final result is obtained to provide a collision prompt, specifically: The final result is sent to the vehicle's IVI or sound system for collision prompting.

6. A parking distance control method including visual collision monitoring as claimed in claim 1, characterized in that: Before the ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle, it also includes: The vehicle speed is obtained, and when the vehicle speed meets the set low-speed range and the vehicle PDC activation conditions are met, the distance information between the vehicle and the obstacle is obtained based on the ultrasonic radar arranged around the vehicle.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the parking distance control method including visual collision monitoring as claimed in any one of claims 1 to 6 are implemented.

8. A parking distance control device including visual collision monitoring, characterized in that: include: Ultrasonic sensing module, which is used to obtain the distance information between the vehicle and the obstacle based on the ultrasonic radar arranged around the vehicle, and obtain the PDC recognition result; A visual perception module, which is used to obtain obstacle information based on the collected images of the vehicle's surrounding environment, and perform collision detection based on the obstacle information and the vehicle's running trajectory; A fusion judgment module is used to weight and fuse the PDC recognition result and the collision detection result to obtain the final result for collision prompt; The ultrasonic radar arranged around the vehicle obtains the distance information between the vehicle and the obstacle and obtains the PDC recognition result. The specific steps include: Based on the coverage of ultrasonic radars arranged around the vehicle, the area around the vehicle is divided into multiple PDC areas; According to the detection of ultrasonic radar, the distance information between the obstacle and the vehicle in each PDC area is obtained to obtain the PDC recognition result; Wherein, the collision detection is performed based on the obstacle information and the vehicle running trajectory, wherein the collision detection method includes a projection method and a triangular area method.

Images