Method and system for testing automatic driving function of intelligent networked automobile
By integrating multiple test items and the YOLO-obb rotating target detection algorithm, a rapid and accurate evaluation of the autonomous driving function of connected vehicles was achieved, solving the problem of low efficiency of manual detection in existing technologies and improving testing efficiency and accuracy.
Patent Information
- Application Number
- CN202511054453.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-30
- Publication Date
- 2025-10-17
AI Technical Summary
In existing technologies, the detection methods for autonomous driving functions require human intervention, making it difficult to quantify the quality of the function. Furthermore, they can only evaluate whether the function is usable, but cannot effectively assess its performance.
By integrating multiple test items and utilizing real-time positioning and image recognition algorithms, the autonomous driving function of connected vehicles is quantified. The YOLO-obb rotating target detection algorithm is used to identify key points on the vehicle body and parking space edge lines, calculate the distance between the wheels and the parking space edge lines, and evaluate the line-crossing situation of the driving function.
It enables rapid and automated testing of connected vehicle driving functions, improving testing efficiency and accuracy, and providing a quantitative assessment of the performance of various driving functions.
Smart Images
Figure CN120800828A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of software testing of connected vehicles, and particularly relates to a connected vehicle automatic driving function testing method and system. BACKGROUND
[0002] With the continuous development of automatic driving vehicles, automobile manufacturers need to detect the related functions of automatic driving offline. The existing solutions are mostly single detection for different functions, and the detection process needs the full-time participation of workers. The workers record the relevant detection data and then calculate the evaluation index according to the provisions in the relevant standard documents to judge whether the function meets the requirements. The existing detection methods are mostly for function development stage detection, and can only evaluate whether the function is usable, but cannot evaluate the quality of the function. SUMMARY
[0003] In view of the deficiencies in the prior art, the present application provides a connected vehicle automatic driving function testing method and system, which quantifies the quality of different functions of the connected vehicle by integrating multiple test items, thereby improving the testing efficiency. The specific technical solutions are as follows: In a first aspect, a connected vehicle automatic driving function testing method is provided. In a first implementation manner of the first aspect, the method comprises: obtaining a real-time position of the connected vehicle, and matching a current test item of the connected vehicle according to the real-time position; obtaining various test parameters corresponding to the test item, and quantifying the use effect of the automatic driving function of the connected vehicle according to all the test parameters.
[0004] In a second implementation manner of the first aspect, the real-time position of the connected vehicle is obtained by: obtaining a real-time image of the connected vehicle, and identifying and positioning the real-time position by using a target detection algorithm.
[0005] In a third implementation manner of the first aspect, the test item comprises: brake test, blind area test, door opening warning test, reverse collision test, side parking test, vertical parking test, and / or rear collision test.
[0006] In a fourth implementation manner of the first aspect, the various test parameters corresponding to the test item are obtained by: obtaining a real-time test image of the connected vehicle during the test, and extracting corresponding test parameters from the real-time test image by using an image recognition algorithm matched with the test item.
[0007] In the fifth implementation manner of the first aspect, in the sixth implementation manner of the first aspect, the test parameters are extracted from the real-time test image, including: The YOLO-obb rotating target detection algorithm is used to identify the body center coordinates, body size and body tilt angle of the connected car in the real-time test image, and the position coordinates of the key points of the parking space boundary line. The distance between each wheel of the connected car and the corresponding parking space boundary line is calculated based on the body center coordinates, body size and body tilt angle, and the position coordinates of the key points.
[0008] In the fifth implementation manner of the first aspect, in the sixth implementation manner of the first aspect, the test parameters are extracted from the real-time test image, including: The YOLO-obb rotating target detection algorithm is used to identify the body center coordinates, body size and body tilt angle of the connected car in the real-time test image, and the position coordinates of the key points of the parking space boundary line. The position coordinates of each endpoint of the connected car body are determined based on the body center coordinates, body size and body tilt angle, and the body pixel point set is determined based on the position coordinates of all endpoints. The parking space boundary line pixel point set is determined based on the position coordinates of the key points of the parking space boundary line. The body pixel point set and the parking space boundary line pixel point set are compared, and the line pressing condition of the connected car in the test process is determined based on the comparison result.
[0009] In the first implementation manner of the second aspect, an intelligent connected car automatic driving function test system is provided, including: A project matching module configured to obtain the real-time position of the connected car, and match the current test project of the connected car based on the real-time position. A function test module configured to obtain each test parameter corresponding to the test project, and quantify the use effect of the automatic driving function of the connected car based on all test parameters.
[0010] In the first implementation manner of the second aspect, in the second implementation manner of the second aspect, the function test module includes: An image recognition unit configured to obtain the real-time test image of the connected car in the test process, and extract the corresponding test parameters from the real-time test image by using an image recognition algorithm matched with the test project.
[0011] In the second implementation manner of the second aspect, in the third implementation manner of the second aspect, the image recognition unit includes: The feature extraction unit is configured to identify the body center coordinates, the body size and the body tilt angle of the connected car and the position coordinates of each key point of the parking space boundary line in the real-time test image by using a YOLO-obb rotating target detection algorithm. The parameter calculation unit is configured to calculate the distance between each wheel of the connected car and the corresponding parking space boundary line in combination with the body center coordinates, the body size and the body tilt angle and the position coordinates of each key point.
[0012] In a fourth implementation manner of the second aspect, in the second implementation manner of the second aspect, the image recognition unit comprises: The feature extraction unit is configured to identify the body center coordinates, the body size and the body tilt angle of the connected car and the position coordinates of each key point of the parking space boundary line in the real-time test image by using a YOLO-obb rotating target detection algorithm. The body pixel extraction unit is configured to determine the position coordinates of each endpoint of the connected car body according to the body center coordinates, the body size and the body tilt angle, and determine a body pixel point set based on the position coordinates of all endpoints. The parking space pixel extraction unit is configured to determine a parking space boundary line pixel point set through the position coordinates of each key point of the parking space boundary line. The parking line pressing ratio unit is configured to compare the body pixel point set with the parking space boundary line pixel point set, and determine the line pressing condition of the connected car in the test process according to the comparison result.
[0013] Beneficial effects: By using the intelligent connected car automatic driving function test method and system, the connected car drives in the test field integrated with multiple test items, the real-time position of the connected car is located, and the test item corresponding to the current position of the connected car can be determined. The test item is used to control the connected car to execute the corresponding test instruction, and each test parameter of the connected car in the test process is collected in real time, and finally the driving function of the connected car in the corresponding test item is quantitatively evaluated by using all the test parameters, so that the driving function of the connected car is quickly tested. BRIEF DESCRIPTION OF DRAWINGS
[0014] In order to more clearly illustrate the specific embodiments of the present application, the drawings required in the specific embodiments will be briefly introduced below. In all the drawings, the elements or parts are not necessarily drawn according to the actual scale.
[0015] Figure 1 The flowchart of the intelligent connected car automatic driving function test method provided by an embodiment of the present application; Figure 2This is a system block diagram of a system for testing the autonomous driving function of an intelligent connected vehicle provided by one embodiment of the present invention; Figure 3 This is a layout diagram of some test scenarios in the test site; Figure 4 for Figure 3 Side view of the test scene shown. DETAILED DESCRIPTION
[0016] The following embodiments of the technical solution of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore only examples and are not intended to limit the scope of protection of the present invention.
[0017] like Figure 1 The flowchart of the method for testing the automatic driving function of an intelligent connected vehicle shown in FIG. 1 includes: Step 1: Obtain the real-time location of the connected vehicle and match the current test item of the connected vehicle based on the real-time location; Step 2: Obtain various test parameters corresponding to the test items, and quantify the effectiveness of the autonomous driving function of the connected vehicle based on all test parameters.
[0018] Specifically, the connected vehicle's real-time location within the test field can be determined. The test field is divided into multiple test areas, each corresponding to different test items. These areas are used to test the connected vehicle's driving functions and evaluate the performance of those functions. By matching the connected vehicle's real-time location with the location ranges of each test area, the test items corresponding to the connected vehicle's current location are determined.
[0019] Then, the connected vehicle can be controlled to execute the corresponding driving actions according to the test instructions corresponding to the test items, and various test parameters of the connected vehicle during the test process can be obtained in real time. These test parameters are matched with the test items, and by analyzing all the obtained test parameters, the driving function of the connected vehicle can be quantitatively evaluated. This enables automated testing of the connected vehicle's driving functions and improves the testing efficiency of the connected vehicle's driving functions.
[0020] In this embodiment, optionally, obtaining the real-time location of the connected vehicle includes: Acquire real-time images of connected vehicles and use target detection algorithms to identify and locate the real-time position.
[0021] Specifically, real-time images of connected vehicles can be collected through cameras deployed in the test field, and existing target detection algorithms, such as the YOLO algorithm, can be used to identify connected vehicles in real-time images and locate the real-time position of connected vehicles.
[0022] In the embodiment, the test items include: a braking test, a blind area test, a door opening warning test, a reversing collision test, a side parking test, a vertical parking test, and / or a rear collision test.
[0023] Specifically, as shown in Figure 3 、 Figure 4 , the test field includes a braking test area for testing the braking function of the connected vehicle, a blind area test area for testing the field of view of the vehicle, a door opening test area for testing the door opening warning function of the connected vehicle, a reversing collision test area for testing the reversing warning function of the vehicle, a side parking test area for testing the side parking function of the vehicle, a vertical parking test area for testing the vertical parking function of the vehicle, and a rear collision test area for testing the rear collision warning function of the vehicle.
[0024] Different test areas are respectively arranged with corresponding test scenes. For example, a moving dummy is arranged on the vehicle driving road in the braking test area. During the test, the moving dummy can cross the driving road at a constant speed according to the test instruction. At the same time, the connected vehicle can collide with the moving dummy at a constant speed according to the test instruction, and start the emergency braking program. Finally, the distance parameter between the parking position of the connected vehicle and the moving dummy is obtained, and the braking performance of the connected vehicle is quantitatively evaluated by the distance parameter.
[0025] For example, a moving dummy car is arranged in the door opening test area. After the connected vehicle moves to the door opening test area, the moving dummy car can drive towards the connected vehicle according to the test instruction. The test personnel in the vehicle can open the door when the moving dummy car passes the connected vehicle, and record the door opening time and the warning time of the connected vehicle. The door opening warning function of the connected vehicle is quantitatively evaluated by the door opening time and the warning time.
[0026] In the embodiment, the test parameters corresponding to the test items are obtained, including: The real-time test image of the connected vehicle during the test is obtained, and the corresponding test parameters are extracted from the real-time test image by using an image recognition algorithm matched with the test item.
[0027] Specifically, for the test items such as the braking test, the reversing collision test, the side parking test, and the vertical parking test, an image recognition algorithm matched with the test item can be used to extract the test parameters for quantitatively evaluating the corresponding driving function of the connected vehicle from the real-time test image obtained during the test.
[0028] For example, for the side parking test, the existing image positioning algorithm can be used to locate the real-time position of the connected car in the side parking video, so as to determine whether there is a line pressing situation in the parking process of the connected car, and the distance between the final parking position and the parking space boundary line, so as to quantitatively evaluate the side parking function of the connected car.
[0029] In this embodiment, optionally, the corresponding test parameters are extracted from the real-time test image, including: The YOLO-obb rotating target detection algorithm is used to identify the body center coordinates, body size and body inclination angle of the connected car in the real-time test image, and the position coordinates of the key points of the parking space boundary line; The distance between each wheel of the connected car and the corresponding parking space boundary line is calculated in combination with the body center coordinates, body size and body inclination angle, and the position coordinates of the key points.
[0030] Specifically, when quantitatively evaluating the side parking or vertical parking function of the connected car, the real-time test image of the parking process of the connected car can be collected in real time by the camera. Then the existing YOLO-obb rotating target detection algorithm can be used to identify the connected car and the parking space boundary line in the real-time test image, and determine the body center coordinates, body size and body inclination angle of the connected car, and the position coordinates of the four endpoints of the parking space boundary line. Then, according to the body center coordinates, body size and body inclination angle, the position coordinates of the four endpoints of the body can be calculated. Through the position coordinates of the four endpoints of the body and the distance between the wheels and the vehicle end, the position coordinates of the four wheel edge contact points of the vehicle can be located.
[0031] The position coordinates of the left front wheel edge contact point are ; ; The position coordinates of the right front wheel edge contact point are ; ; The position coordinates of the left rear wheel edge contact point are ; ; The position coordinates of the right rear wheel edge contact point are ; ; wherein , , , are the position coordinates of the left front end point, the right front end point, the left rear end point and the right rear end point of the body, respectively. , respectively represent the distance between the front wheel and the front end of the vehicle, the distance between the rear wheel and the rear end of the vehicle, respectively represent the height of the vehicle body.
[0032] The position coordinates of the corresponding outer edge grounding points and the position coordinates of the key points of the corresponding parking space edge lines are combined to calculate the distance between each wheel and the corresponding parking space edge line.
[0033] Taking the distance between the left front wheel of the vehicle and the left edge line of the parking space as an example. First, the position coordinates of the left front end point of the parking space frame line and the position coordinates of the left rear end point are used to create a straight line, and the expression of the straight line is: ; Then, the expression is transformed into the form of , and the constant , , , , can be obtained according to the known position coordinates
[0034] Finally, the outer edge grounding point of the left front wheel is combined to calculate the distance between the left front wheel and the left edge line of the parking space, and the specific calculation formula is: ; wherein, is the length of the parking space, .
[0035] The calculation methods of the distances between the right front wheel, the left rear wheel, the right rear wheel and the parking space edge line are the same as the algorithm of the distance between the left front wheel and the left edge line of the parking space, which will not be described again.
[0036] In this embodiment, the corresponding test parameters can be extracted from the real-time test image, including: The YOLO-obb rotating target detection algorithm is used to identify the body center coordinates, body size and body inclination angle of the connected car in the real-time test image, and the position coordinates of each key point of the parking space edge line; According to the body center coordinates, the body size and the body inclination angle, the position coordinates of each end point of the connected car body are determined, and the body pixel point set is determined based on the position coordinates of all end points; The position coordinates of each key point of the parking space edge line are used to determine the parking space edge line pixel point set; The vehicle body pixel point set is compared with the parking space boundary line pixel point set, and the line pressing condition of the connected car in the test process is determined according to the comparison result.
[0037] Specifically, before calculating the distance between the wheel and the parking space boundary line, it can be determined in advance whether the wheel exists in the line pressing condition, and if so, the distance between the wheel and the parking space boundary line does not need to be calculated, thereby improving the test efficiency. At the same time, the line pressing condition of the connected car during parking can be monitored in real time to evaluate the parking performance of the connected car.
[0038] Specifically, first, the existing YOLO-obb rotating target detection algorithm can be used to identify and locate the vehicle body center coordinates, vehicle body size and vehicle body inclination angle of the connected car in the real-time test image, as well as the position coordinates of the key points of the parking space boundary line. Then, the position coordinates of the four end points of the vehicle body can be calculated according to the vehicle body center coordinates, vehicle body size and vehicle body inclination angle. Based on the position coordinates of the four end points, the pixel point range covered by the vehicle body is determined to obtain the vehicle body pixel point set in the real-time test image. Similarly, according to the position coordinates of the 4 key points of the parking line boundary line, the parking space boundary line pixel point set in the real-time test image can be obtained. Finally, the parking space boundary line pixel point set and the vehicle body pixel point set are compared. If the vehicle body pixel point set includes the elements in the parking space boundary line pixel point set, it indicates that the connected car in the real-time test image exists in the line pressing condition. Otherwise, it indicates that the connected car in the real-time test image does not exist in the line pressing condition, and the distance between the wheel and the parking space boundary line can be calculated using the above algorithm.
[0039] As shown in the system block diagram of the intelligent connected car automatic driving function test system, the test system comprises: Figure 2 A project matching module configured to obtain the real-time position of the connected car and match the current test project of the connected car according to the real-time position; A function test module configured to obtain each test parameter corresponding to the test project, and to quantify the use effect of the automatic driving function of the connected car according to all the test parameters. Specifically, the test system comprises a project matching module and a function test module. The project matching module can locate the real-time position of the connected car in the test field in real time. The test field is divided into a plurality of test areas, each of which corresponds to a different test project, for testing the corresponding driving function of the connected car to evaluate the driving function of the connected car. By matching the real-time position of the connected car with the position range of each test area, the test project corresponding to the position area where the connected car is currently located is determined.
[0040]
[0041] The function test module can control the connected car to perform corresponding driving actions according to the test instruction corresponding to the test item, and can obtain various test parameters of the connected car in the test process in real time. The test parameters match the test item, and the driving function of the connected car can be quantitatively evaluated by analyzing all the obtained test parameters. Thus, the automatic test of the driving function of the connected car is realized, and the test efficiency of the driving function of the connected car is improved.
[0042] In this embodiment, the function test module comprises, optionally: An image recognition unit configured to obtain real-time test images of the connected car in the test process, and extract corresponding test parameters from the real-time test images by using an image recognition algorithm matched with the test item.
[0043] Specifically, the function test module comprises an image recognition unit, which can collect real-time images of the connected car through a camera arranged in the test field, and identify the connected car in the real-time images by using an existing target detection algorithm such as YOLO algorithm, and locate the real-time position of the connected car.
[0044] In this embodiment, the image recognition unit comprises, optionally: A feature extraction unit configured to identify the center coordinates of the vehicle body, the size of the vehicle body, and the tilt angle of the vehicle body of the connected car in the real-time test images, and the position coordinates of the key points of the edge line of the parking space by using a YOLO-obb rotating target detection algorithm; A parameter calculation unit configured to calculate the distance between each wheel of the connected car and the corresponding edge line of the parking space in combination with the center coordinates of the vehicle body, the size of the vehicle body, and the tilt angle of the vehicle body, and the position coordinates of the key points.
[0045] Specifically, the image recognition unit comprises a feature extraction unit and a parameter calculation unit. The feature extraction unit can identify the connected car and the edge line of the parking space in the real-time test images by using an existing YOLO-obb rotating target detection algorithm, and determine the center coordinates of the vehicle body, the size of the vehicle body, and the tilt angle of the vehicle body of the connected car, and the position coordinates of the four end points of the edge line of the parking space. The parameter calculation unit can calculate the position coordinates of the four end points of the vehicle body according to the center coordinates of the vehicle body, the size of the vehicle body, and the tilt angle of the vehicle body. The position coordinates of the four wheel edge contact points of the vehicle can be located by the position coordinates of the four end points of the vehicle body and the distance between the wheels and the end of the vehicle. In combination with the position coordinates of the corresponding wheel edge contact points and the position coordinates of the key points of the corresponding edge line of the corresponding parking space, the distance between each wheel and the corresponding edge line of the parking space can be calculated.
[0046] In this embodiment, the image recognition unit comprises, optionally: The feature extraction unit is configured to identify the body center coordinates, the body size and the body tilt angle of the connected car in the real-time test image and the position coordinates of the key points of the parking space boundary line by using a YOLO-obb rotating target detection algorithm. The body pixel extraction unit is configured to determine the position coordinates of each endpoint of the connected car body according to the body center coordinates, the body size and the body tilt angle, and determine the body pixel set based on the position coordinates of all the endpoints. The parking space pixel extraction unit is configured to determine the parking space boundary line pixel set through the position coordinates of the key points of the parking space boundary line. The parking line pressing ratio unit is configured to compare the body pixel set with the parking space boundary line pixel set, and determine the pressing line condition of the connected car in the test process according to the comparison result.
[0047] Specifically, the image recognition unit not only includes the feature extraction unit and the parameter calculation unit, but also includes the body pixel extraction unit, the parking space pixel extraction unit and the parking line pressing comparison unit. The body pixel extraction unit can calculate the position coordinates of the four endpoints of the body according to the body center coordinates, the body size and the body tilt angle. Based on the position coordinates of the four endpoints, the pixel point range covered by the body is determined to obtain the body pixel set in the real-time test image. Similarly, the parking space pixel extraction unit can obtain the parking space boundary line pixel set in the real-time test image according to the position coordinates of the four key points of the parking line boundary line. The parking line pressing comparison unit can compare the parking space boundary line pixel set with the body pixel set. If the body pixel set includes the elements in the parking space boundary line pixel set, it indicates that the connected car in the real-time test image has the pressing line condition. Otherwise, it indicates that the connected car in the real-time test image does not have the pressing line condition, and the distance between the wheel and the parking space boundary line can be calculated by using the above algorithm.
[0048] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit it; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and they should be covered in the scope of the claims and the specification of the present application.
Claims
1. A method for testing the automatic driving function of an intelligent connected vehicle, characterized in that: include: Obtain the real-time location of the connected vehicle and match the current test items of the connected vehicle based on the real-time location; Obtain various test parameters corresponding to the test items, and quantify the effectiveness of the autonomous driving function of the connected vehicle based on all test parameters.
2. The method for testing the automatic driving function of an intelligent connected vehicle according to claim 1, characterized in that: Get the real-time location of connected cars, including: Acquire real-time images of connected vehicles and use target detection algorithms to identify and locate the real-time position.
3. The method for testing the automatic driving function of an intelligent connected vehicle according to claim 1, characterized in that: The test items include: braking test, blind spot test, door opening warning test, reversing collision test, parallel parking test, perpendicular parking test and / or rear collision test.
4. The method for testing the automatic driving function of an intelligent connected vehicle according to claim 1, wherein: Obtain the test parameters corresponding to the test item, including: Acquire a real-time test image of the connected vehicle during the test process, and extract corresponding test parameters from the real-time test image using an image recognition algorithm that matches the test item.
5. The method for testing the automatic driving function of an intelligent connected vehicle according to claim 4, characterized in that: Extracting corresponding test parameters from the real-time test image includes: Using the YOLO-obb rotating target detection algorithm to identify the body center coordinates, body size, and body tilt angle of the connected vehicle in the real-time test image, as well as the position coordinates of key points on the parking space boundary line; The distance between each wheel of the connected vehicle and the edge of the corresponding parking space is calculated by combining the vehicle body center coordinates, vehicle body size and vehicle body tilt angle, as well as the position coordinates of each key point.
6. The method for testing the automatic driving function of an intelligent connected vehicle according to claim 4, characterized in that: Extracting corresponding test parameters from the real-time test image includes: Using the YOLO-obb rotating target detection algorithm to identify the body center coordinates, body size, and body tilt angle of the connected vehicle in the real-time test image, as well as the position coordinates of key points on the parking space boundary line; Determine the position coordinates of each endpoint of the connected vehicle body based on the body center coordinates, body size, and body tilt angle, and determine a body pixel point set based on the position coordinates of all endpoints; Determine a parking space boundary pixel point set based on the position coordinates of each key point of the parking space boundary; The vehicle body pixel point set is compared with the parking space edge pixel point set, and the line crossing situation of the connected vehicle during the test is determined based on the comparison result.
7. A system for testing the automatic driving function of an intelligent connected vehicle, characterized in that: include: An item matching module is configured to obtain the real-time location of the connected vehicle and match the current test item of the connected vehicle according to the real-time location; The functional testing module is configured to obtain various test parameters corresponding to the test items and quantify the effectiveness of the automatic driving function of the connected vehicle based on all the test parameters.
8. The intelligent connected vehicle automatic driving function test system according to claim 7, characterized in that: The functional testing module includes: The image recognition unit is configured to obtain a real-time test image of the connected vehicle during the test process, and extract corresponding test parameters from the real-time test image using an image recognition algorithm that matches the test item.
9. The intelligent connected vehicle automatic driving function test system according to claim 8, characterized in that: The image recognition unit includes: a feature extraction unit configured to use a YOLO-obb rotating target detection algorithm to identify the body center coordinates, body size, and body tilt angle of the connected vehicle in the real-time test image, as well as the position coordinates of key points of the parking space boundary line; The parameter calculation unit is configured to calculate the distance between each wheel of the connected vehicle and the corresponding parking space edge line based on the vehicle body center coordinates, vehicle body size and vehicle body tilt angle, as well as the position coordinates of each key point.
10. The intelligent connected vehicle automatic driving function test system according to claim 8, characterized in that: The image recognition unit includes: a feature extraction unit configured to use a YOLO-obb rotating target detection algorithm to identify the body center coordinates, body size, and body tilt angle of the connected vehicle in the real-time test image, as well as the position coordinates of key points of the parking space boundary line; a body pixel extraction unit configured to determine the position coordinates of each endpoint of the connected vehicle body based on the body center coordinates, body size, and body tilt angle, and determine a body pixel point set based on the position coordinates of all endpoints; a parking space pixel extraction unit configured to determine a parking space boundary pixel point set based on the position coordinates of each key point of the parking space boundary; The parking line pressing ratio unit is configured to compare the vehicle body pixel point set with the parking space edge pixel point set, and determine the line pressing situation of the connected vehicle during the test according to the comparison result.