Automatic testing system and method for AI-assisted functions of electronic rearview mirrors

Abstract

The present invention discloses an automated testing system and method for the AI ​​auxiliary function of an electronic rearview mirror, and belongs to the technical field of advanced driver assistance systems. The automated testing system for the AI ​​auxiliary function of an electronic rearview mirror includes a vehicle body signal module, a server, a CMS electronic rearview mirror controller, a left camera, a right camera, a left display screen, a right display screen, and a display; the server is used to simulate the sending of CAN signals, and verify the consistency of the icons and signal values ​​in the left and right display screens, and the distance display values ​​in the log files; the CMS electronic rearview mirror controller is used to control the left and right display screens to display environmental image information, and perform signal detection processing and early warning control. The present invention provides an automated testing system and method for the AI ​​auxiliary function of an electronic rearview mirror, which realizes the automated testing of the auxiliary function of the CMS electronic rearview mirror, reduces labor costs, improves testing efficiency, and can also simulate long-term stable testing, which is conducive to ensuring the stability of the auxiliary function.

Description

Technical Field

[0001] The present invention relates to an automatic testing system and method for an AI auxiliary function of an electronic rearview mirror, and belongs to the technical field of advanced driver assistance systems. Background Art

[0002] At present, with the continuous development of vehicle intelligence, more and more driving assistance functions are being applied to real vehicles. With the official promulgation of domestic regulations on automotive electronic rearview mirrors, more and more cars are beginning to install CMS electronic rearview mirrors to replace traditional physical rearview mirrors.

[0003] In the existing technology, when testing the actual vehicle, it is necessary to rely on the human eye to compare the image display to determine whether the actual result of the CMS electronic rearview mirror auxiliary function is consistent with the expected result. Alternatively, when the collected data is refilled and verified on the test bench, the human eye is still required to determine whether the actual result in the refill test is consistent with the expected result. When conducting a long-term auxiliary function test on a real vehicle, due to the high frame rate of the video on the display screen, it is easy to cause problems such as missed detection or untimely recording due to human factors. In addition, when a problem is found and the scene needs to be reproduced, it is impossible to obtain the required scene in time for positioning analysis or refill test verification, affecting the test efficiency. When refilling the test on the test bench using road data, only the warning function of the target area can be verified, and the verification of auxiliary functions such as field of view switching and DOW door opening warning cannot be covered. In addition, the human eye is also required to judge the output results in the refill test, which will also cause problems such as missed detection or untimely recording. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide an automated testing system and method for the AI ​​auxiliary function of an electronic rearview mirror, which realizes the automated testing of the auxiliary function of a CMS electronic rearview mirror, can timely obtain problem scenarios during the test process and perform positioning analysis and repeated verification tests, thereby reducing the manpower cost of development and verification, shortening the test time, and improving the test efficiency. At the same time, a large amount of data can also be repeatedly verified to simulate long-term stable testing, which is conducive to ensuring the stability of the auxiliary function.

[0005] In order to solve the above technical problems, the technical solution of the present invention is:

[0006] On one hand, the present invention provides an automated test system for AI auxiliary functions of an electronic rearview mirror, which includes a vehicle body signal module, a server, a CMS electronic rearview mirror controller, a left camera, a right camera, a left display screen, a right display screen, and a display;

[0007] The server is used to receive the body signal sent by the body signal module, simulate the sending of the CAN signal, verify the consistency between the icons in the left and right display screens and the signal value, and verify the consistency between the icons in the left and right display screens and the distance display in the log file;

[0008] The left camera and the right camera are used to collect image information of the environment behind the vehicle;

[0009] The CMS electronic rearview mirror controller is used to receive the CAN signal sent by the server, receive the vehicle side rear environment image information collected by the left camera and the right camera, control the left display screen and the right display screen to display the vehicle side rear environment image information, and perform signal detection processing and early warning control;

[0010] The display is used to play the source video captured by the camera;

[0011] The left display screen and the right display screen are used to display the external environment information after image processing by the CMS electronic rearview mirror controller.

[0012] Further, the vehicle body signal module includes an ESP module, a TCU module, a BCM module and a CEM module;

[0013] The ESP module is used to obtain vehicle speed information;

[0014] The TCU module is used to obtain a gear position signal;

[0015] The BCM module is used to obtain the door switch signal;

[0016] The CEM module is used to obtain a turn indication signal.

[0017] Another aspect of the present invention further provides a testing method of an AI auxiliary function automated testing system for an electronic rearview mirror, which comprises the following steps:

[0018] Step S1, the recording function of the CMS electronic rearview mirror controller is set to record the Vo output video;

[0019] Step S2, turning on the switch for recording the log file in the CMS electronic rearview mirror controller, and the CMS electronic rearview mirror controller saves the human-vehicle recognition results of the images captured by the left camera and the right camera;

[0020] Step S3, the CMS electronic rearview mirror controller records the source video captured by the left camera during the driving process of the vehicle, and plays it through the display, then aims the left camera at the display to shoot, adjusts the positions of the left camera and the display, and observes the BSD icon, distance display icon and DOW icon displayed on the left display screen;

[0021] Step S4, the first Python script reads and records the csv file of the CAN message, and sends the vehicle speed, gear position, and turn indication signal to the CMS electronic rearview mirror controller according to the message format period in the DBC file;

[0022] Step S5, a second Python script synchronously controls the video playback of step S3 and the signal injection of step S4;

[0023] Step S6, process the Vo output video recorded in step S1, obtain each frame of the picture and save it to the server directory;

[0024] Step S7: Execute the Python test script to perform auxiliary function test.

[0025] Further, in step S7, executing the Python test script to perform auxiliary function testing specifically includes the following steps:

[0026] Execute the first Python test script to perform the field of view switching test;

[0027] Execute the second Python test script to perform BSD testing;

[0028] Execute the third Python test script to perform DOW test;

[0029] Execute the fourth Python test script to perform view switching, BSD and DOW batch testing;

[0030] Execute the fifth Python test script to perform field of view switching, BSD and DOW long-term stability tests.

[0031] Further, executing the first Python test script to perform the field of view switching test specifically includes the following steps:

[0032] Step S711: The first Python test script parses the CAN message to obtain the gear position signal value and the left turn indication signal value, and when the gear position is R and the left turn indication signal value is Active, a first timestamp is recorded;

[0033] Step S712: search the pictures in the server directory according to the first timestamp, and compare them with the preset reverse and left turn icons to find out whether the reverse icon and the left turn icon exist in the server directory. If they exist, the test passes; if not, jump to step S713;

[0034] Step S713: If the test fails, record the picture frame number.

[0035] Further, executing the second Python test script to perform the BSD test specifically includes the following steps:

[0036] Step S721, the second Python test script parses the log file to determine whether the person and the vehicle meet the lateral and longitudinal distance conditions. If the conditions are met, jump to step S722; if the conditions are not met, repeat step S721;

[0037] Step S722, recording a second timestamp when the condition is met;

[0038] Step S723, traverse and search for pictures in the server directory according to the second timestamp, and compare with the preset BSD and distance display icon to find out whether the BSD and distance display icon exist in the server directory. If they exist, the test passes; if not, jump to step S724;

[0039] Step S724: If the test fails, record the picture frame number.

[0040] Further, the execution of the third Python test script to perform the DOW test specifically includes the following steps:

[0041] Step S731, the third Python test script parses the log file to determine whether the person and the vehicle meet the lateral and longitudinal distance conditions and the relative speed condition with the vehicle. If the conditions are met, jump to step S732; if the conditions are not met, repeat step S731;

[0042] Step S732, recording a third timestamp when the condition is met;

[0043] Step S733, parsing the vehicle speed signal value in the CAN message according to the third timestamp, if the vehicle speed condition is met, jumping to step S734, if the vehicle speed condition is not met, repeating step S733;

[0044] Step S734, traverse and search for pictures in the server directory according to the third timestamp, and compare with the preset DOW icon to find out whether the DOW icon exists in the server directory. If so, the test passes; if not, jump to step S735;

[0045] Step S735: If the test fails, record the picture frame number.

[0046] Further, the execution of the fourth Python test script to perform field of view switching, BSD and DOW batch testing specifically includes the following steps:

[0047] Step S741, storing the csv files of the source video data and CAN messages collected at the same time point in a server folder;

[0048] Step S742, traverse the data in each folder for testing, and execute the test in the order of view switching, BSD and DOW until the data in all folders are tested, then exit the test.

[0049] Further, the execution of the fifth Python test script to perform field of view switching, BSD and DOW long-term stability tests specifically includes the following steps:

[0050] Step S751, storing the source video data and CAN message csv files collected at the same time point in a server folder;

[0051] Step S752, traverse the data in each folder to perform a test, and perform the test in the order of view switching, BSD and DOW until the data in all folders are tested;

[0052] Step S753, perform the next loop traversal and test until the shortcut exits.

[0053] By adopting the above technical scheme, the present invention sets a CMS electronic rearview mirror controller to record the video displayed on the left and right display screens, and uses a Python script to parse and obtain each frame of the picture, which is convenient for comparing icons of different tests. By setting a server to receive the body signal sent by the body signal module, it is possible to simulate the sending of CAN signals, and verify the consistency of the icons and signal values ​​and distance display values ​​in the left and right display screens, thereby avoiding the problems of missed detection and untimely recording caused by human eye judgment, improving the accuracy of the test, and reducing labor costs. At the same time, by using a Python script to execute the test and record the frame number of the picture that failed the test, the automatic test of the auxiliary function of the CMS electronic rearview mirror can be realized, and the problem scenarios in the test process can be obtained in time to perform positioning analysis and repeated verification tests, thereby shortening the test time and improving the test efficiency. By using a Python script to perform a loop traversal test, a long-term stable test can be simulated, and a large amount of data can be repeatedly verified, which is beneficial to ensuring the stability of the auxiliary function. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 The principle block diagram of the AI ​​auxiliary function automatic test system of the electronic rearview mirror of the present invention;

[0055] Figure 2 The schematic diagram of the AI ​​auxiliary function automated testing system of the electronic rearview mirror of the present invention;

[0056] Figure 3 A flow chart of a test method of an automated test system for AI auxiliary functions of an electronic rearview mirror of the present invention;

[0057] Figure 4 This is a flow chart of the field of view switching test of the present invention;

[0058] Figure 5 It is the BSD test flow chart of the present invention;

[0059] Figure 6 It is a DOW test flow chart of the present invention. DETAILED DESCRIPTION

[0060] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0061] Embodiment 1

[0062] like Figure 1 As shown, this embodiment provides an AI auxiliary function automated test system for an electronic rearview mirror, which includes a vehicle body signal module, a server, a CMS electronic rearview mirror controller, a left camera, a right camera, a left display screen, a right display screen, and a display;

[0063] The server is used to receive the body signal sent by the body signal module, simulate the sending of CAN signals, verify the consistency between the icons on the left and right display screens and the signal values, and verify the consistency between the icons on the left and right display screens and the distance display in the log file;

[0064] The left camera and the right camera are used to collect image information of the vehicle's rear side environment, replacing the original physical rearview mirror;

[0065] The CMS electronic rearview mirror controller is used to receive CAN signals sent by the server, receive the vehicle side rear environment image information collected by the left camera and the right camera, control the left display screen and the right display screen to display the vehicle side rear environment image information, and perform signal detection processing and early warning control;

[0066] The display is used to play the source video captured by the camera;

[0067] The left display screen and the right display screen are used to display the external environment information after image processing by the CMS electronic rearview mirror controller.

[0068] like Figure 2 As shown, the vehicle body signal module of this embodiment includes an ESP module, a TCU module, a BCM module and a CEM module;

[0069] The ESP module is a vehicle stability system used to obtain vehicle speed information;

[0070] The TCU module is an automatic transmission control unit, used to obtain gear position signals;

[0071] The BCM module is a body controller used to obtain door switch signals;

[0072] The CEM module is the central electronic module, which is used to obtain turn indication signals.

[0073] Embodiment 2

[0074] like Figure 3 As shown, this embodiment provides a test method of the AI ​​auxiliary function automatic test system of the electronic rearview mirror as in Embodiment 1, which comprises the following steps:

[0075] Step S1, the recording function of the CMS electronic rearview mirror controller is set to record the Vo output video, that is, to record the video displayed on the left display screen and the right display screen;

[0076] Step S2: Turn on the switch for recording the dumpframe log file in the CMS electronic rearview mirror controller, and the CMS electronic rearview mirror controller saves the human-vehicle recognition results of the images captured by the left camera and the right camera;

[0077] Step S3, the CMS electronic rearview mirror controller records the source video captured by the left camera during the driving process of the vehicle, and plays it through the display, then aims the left camera at the display to shoot, adjusts the positions of the left camera and the display, and observes the BSD icon, distance display icon and DOW icon displayed on the left display screen;

[0078] Step S4, the first Python script reads and records the csv file of the CAN message, and sends the vehicle speed, gear position, and turn indication signal to the CMS electronic rearview mirror controller according to the message format period in the DBC file (CAN communication matrix), so as to achieve the purpose of simulating communication with the whole vehicle during the actual vehicle test;

[0079] Step S5, the second Python script synchronously controls the video playback of step S3 and the signal injection of step S4 to ensure the first frame synchronization of the video signal and the CAN signal;

[0080] Step S6, process the Vo output video recorded in step S1, obtain each frame of the image and save it to the vo_video_output directory;

[0081] Step S7: Execute the Python test script to perform auxiliary function test.

[0082] In step S7 of this embodiment, executing the Python test script to perform auxiliary function testing specifically includes the following steps:

[0083] Execute the first Python test script python cmstest.py viewtest to perform the view switching test;

[0084] Execute the second Python test script python cmstest.py bsdtest to perform BSD test;

[0085] Execute the third Python test script python cmstest.py dowtest to perform DOW test;

[0086] Execute the fourth Python test script python cmstest.py / testdir to perform view switching, BSD and DOW batch testing;

[0087] Execute the fifth Python test script python cmstest.py / testdir-w to perform view switching, BSD and DOW long-term stability tests.

[0088] like Figure 4 As shown, in this embodiment, the first Python test script python cmstest.py viewtest is executed to perform a view switching test, which specifically includes the following steps:

[0089] Step S711: The first Python test script parses the csv file of the CAN message to obtain the gear position signal value and the left turn indication signal value, and when the gear position is R and the left turn indication signal value is Active, a first timestamp is recorded;

[0090] Step S712, traverse and search for pictures in the vo_video_output directory according to the first timestamp, and compare with the preset reverse and left turn icons to find out whether there is a reverse icon and a left turn icon in the server directory. If they exist, it means that the reverse view switching and the left turn view switching are performed at this time, and the test passes. If they do not exist, jump to step S713;

[0091] Step S713: If the test fails, the picture frame number is recorded to facilitate finding the reason why the field of view is not switched.

[0092] like Figure 5 As shown, in this embodiment, executing the second Python test script python cmstest.py bsdtest to perform BSD testing specifically includes the following steps:

[0093] Step S721: The second Python test script parses the dumpframe log file to determine whether the person and the vehicle meet the following lateral and longitudinal distance conditions:

[0094] The horizontal distance is 0~3m, and the vertical distance is 0~6m (the BSD icon is red), or

[0095] The horizontal distance is 0~3m, and the vertical distance is 6~12m (the BSD icon is yellow);

[0096] If the condition is met, jump to step S722, if the condition is not met, repeat step S721;

[0097] Step S722, recording a second timestamp when the condition is met;

[0098] Step S723, traverse and search for pictures in the vo_video_output directory according to the second timestamp, and compare with the preset BSD and distance display icons to find out whether the BSD and distance display icons exist in the server directory. If they exist, it means that a BSD warning is performed at this time and the test passes. If they do not exist, jump to step S724;

[0099] Step S724: If the test fails, the picture frame number is recorded to facilitate finding the reason why the BSD warning was not issued.

[0100] like Figure 6 As shown, the third Python test script python cmstest.py dowtest in this embodiment is executed to perform a DOW test, which specifically includes the following steps:

[0101] Step S731: The third Python test script parses the dumpframe log file to determine whether the person and the vehicle meet the following lateral and longitudinal distance conditions and relative speed conditions with the vehicle:

[0102] The horizontal distance is -0.5~2m, the vertical distance is 0~15m, and

[0103] The relative speed with the vehicle is ≥ 0.5m / s;

[0104] If the condition is met, jump to step S732, if the condition is not met, repeat step S731;

[0105] Step S732, recording a third timestamp when the condition is met;

[0106] Step S733: parse the vehicle speed signal value in the CAN message according to the third timestamp to determine whether the following vehicle speed conditions are met:

[0107] 0≤Speed≤3km / h;

[0108] If the vehicle speed condition is met, then jump to step S734; if the vehicle speed condition is not met, then repeat step S733;

[0109] Step S734, traverse and search for pictures in the vo_video_output directory according to the third timestamp, and compare with the preset DOW icon to find out whether there is a DOW icon in the server directory. If there is, it means that a DOW warning is issued at this time and the test passes. If not, jump to step S735;

[0110] Step S735: If the test fails, the picture frame number is recorded to facilitate finding the reason why the DOW warning was not issued.

[0111] The fourth Python test script python cmstest.py / testdir of this embodiment is executed to perform view switching and BSD and DOW batch testing, specifically including the following steps:

[0112] Step S741, storing the csv files of the source video data and CAN messages collected at the same time point in a server folder;

[0113] Step S742, traverse the data in each folder in testdir for testing, and execute the test in the order of view switching, BSD and DOW until the data in all folders are tested, then exit the test.

[0114] The fifth Python test script python cmstest.py / testdir-w in this embodiment is executed to perform field of view switching, BSD and DOW long-term stability tests, specifically including the following steps:

[0115] Step S751, storing the source video data and CAN message csv files collected at the same time point in a server folder;

[0116] Step S752, traverse the data in each folder in testdir to perform the test, and execute the test in the order of view switching, BSD and DOW until the data in all folders are tested;

[0117] Step S753: perform the next loop traversal and test until exiting by pressing the ctrl+c shortcut.

[0118] The working principle of the present invention is as follows:

[0119] The Python script is used to synchronously control the vehicle body CAN signal injection and the display's playback of the video recorded by the CMS electronic rearview mirror controller. The CMS electronic rearview mirror controller records the videos displayed on the left and right display screens and parses them. After obtaining each frame of the image, it is compared with the vehicle body CAN signal value at the same time or with the recognition result in the dumpframe file at the same time. It is verified whether the icon display in each frame of the image output by the left and right display screens is correct, and it is automatically determined whether the actual results of the field of view switching function, DOW function, and BSD function are consistent with the expected results.

[0120] By setting up the CMS electronic rearview mirror controller to record the video displayed on the left and right display screens, and using Python scripts to parse and obtain each frame of the image, it is convenient to compare icons for different tests. By setting up the server to receive the body signal sent by the body signal module, it is possible to simulate the sending of CAN signals and verify the consistency of the icons and signal values ​​and distance display values ​​in the left and right display screens, avoiding the problems of missed detection and untimely recording caused by human eye judgment, improving the accuracy of the test, and reducing labor costs. At the same time, using Python scripts to execute tests and record the frame numbers of pictures that failed the tests, it is possible to achieve automated testing of the auxiliary functions of the CMS electronic rearview mirror, and to promptly obtain problem scenarios during the test process, perform positioning analysis, and repeat verification tests, shortening the test time and improving test efficiency. Using Python scripts for loop traversal tests can simulate long-term stable tests and repeat verification of large amounts of data, which is conducive to ensuring the stability of the auxiliary functions.

[0121] The specific embodiments described above further illustrate the technical problems, technical solutions and beneficial effects solved by the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An AI auxiliary function automated testing system for a CMS electronic rearview mirror, characterized in that: It includes a body signal module, a server, a CMS electronic rearview mirror controller, a left camera, a right camera, a left display screen, a right display screen and a display; The server is used to receive the body signal sent by the body signal module, simulate the sending of the CAN signal, verify the consistency between the icons in the left and right display screens and the signal value, and verify the consistency between the icons in the left and right display screens and the distance display in the log file; The left camera and the right camera are used to collect image information of the environment behind the vehicle; The CMS electronic rearview mirror controller is used to receive the CAN signal sent by the server, receive the vehicle side rear environment image information collected by the left camera and the right camera, control the left display screen and the right display screen to display the vehicle side rear environment image information, and perform signal detection processing and early warning control; The display is used to play the source video captured by the camera; The left display screen and the right display screen are used to display the external environment information processed by the CMS electronic rearview mirror controller; The testing method of the AI ​​auxiliary function automatic testing system of the CMS electronic rearview mirror comprises the following steps: Step S1, the recording function of the CMS electronic rearview mirror controller is set to record the Vo output video; Step S2, turning on the switch for recording the log file in the CMS electronic rearview mirror controller, and the CMS electronic rearview mirror controller saves the human-vehicle recognition results of the images captured by the left camera and the right camera; Step S3, the CMS electronic rearview mirror controller records the source video captured by the left camera during the driving process of the vehicle, and plays it through the display, then aims the left camera at the display to shoot, adjusts the positions of the left camera and the display, and observes the BSD icon, distance display icon and DOW icon displayed on the left display screen; Step S4, the first Python script reads and records the csv file of the CAN message, and sends the vehicle speed, gear position, and turn indication signal to the CMS electronic rearview mirror controller according to the message format period in the DBC file; Step S5, a second Python script synchronously controls the video playback of step S3 and the signal injection of step S4; Step S6, process the Vo output video recorded in step S1, obtain each frame of the picture and save it to the server directory; Step S7, executing the Python test script to perform auxiliary function test, including: using the Python script to parse and obtain each frame of the picture, so as to facilitate the comparison of icons for different tests, setting the server to receive the body signal sent by the body signal module, simulating the sending of CAN signals, and verifying the consistency between the icons and the signal values ​​and the distance display values ​​in the left and right display screens, and at the same time, using the Python script to execute the test and record the frame number of the picture that failed the test; the step S7 specifically includes the following steps: Execute the first Python test script to perform the field of view switching test; Execute the second Python test script to perform BSD testing; Execute the third Python test script to perform DOW test; Execute the fourth Python test script to perform view switching, BSD and DOW batch testing; Execute the fifth Python test script to perform field of view switching, BSD and DOW long-term stability tests.

2. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1 is characterized in that: The vehicle body signal module includes an ESP module, a TCU module, a BCM module and a CEM module; The ESP module is used to obtain vehicle speed information; The TCU module is used to obtain a gear position signal; The BCM module is used to obtain the door switch signal; The CEM module is used to obtain a turn indication signal.

3. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1 is characterized in that: The executing of the first Python test script to perform the field of view switching test specifically includes the following steps: Step S711: The first Python test script parses the CAN message to obtain the gear position signal value and the left turn indication signal value, and when the gear position is R and the left turn indication signal value is Active, a first timestamp is recorded; Step S712, traverse and search for pictures in the server directory according to the first timestamp, and compare with the preset reverse and left turn icons to find out whether there are reverse icons and left turn icons in the server directory. If they exist, the test passes; if not, jump to step S713; Step S713: If the test fails, record the picture frame number.

4. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1, characterized in that: The execution of the second Python test script to perform the BSD test specifically includes the following steps: Step S721, the second Python test script parses the log file to determine whether the person and the vehicle meet the lateral and longitudinal distance conditions. If the conditions are met, jump to step S722; if the conditions are not met, repeat step S721; Step S722, recording a second timestamp when the condition is met; Step S723, traverse and search for pictures in the server directory according to the second timestamp, and compare with the preset BSD and distance display icon to find out whether the BSD and distance display icon exist in the server directory. If they exist, the test passes; if not, jump to step S724; Step S724: If the test fails, record the picture frame number.

5. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1 is characterized in that: The execution of the third Python test script to perform the DOW test specifically includes the following steps: Step S731, the third Python test script parses the log file to determine whether the person and the vehicle meet the lateral and longitudinal distance conditions and the relative speed condition with the vehicle. If the conditions are met, jump to step S732; if the conditions are not met, repeat step S731; Step S732, recording a third timestamp when the condition is met; Step S733, parsing the vehicle speed signal value in the CAN message according to the third timestamp, if the vehicle speed condition is met, jumping to step S734, if the vehicle speed condition is not met, repeating step S733; Step S734, traverse and search for pictures in the server directory according to the third timestamp, and compare with the preset DOW icon to find out whether the DOW icon exists in the server directory. If so, the test passes; if not, jump to step S735; Step S735: If the test fails, record the picture frame number.

6. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1, characterized in that: The execution of the fourth Python test script to perform field of view switching, BSD and DOW batch testing specifically includes the following steps: Step S741, storing the csv files of the source video data and CAN messages collected at the same time point in a server folder; Step S742, traverse the data in each folder for testing, and execute the test in the order of view switching, BSD and DOW until the data in all folders are tested, then exit the test.

7. The AI ​​auxiliary function automated testing system of the CMS electronic rearview mirror according to claim 1, characterized in that: The execution of the fifth Python test script to perform field of view switching, BSD and DOW long-term stability tests specifically includes the following steps: Step S751, storing the source video data and CAN message csv files collected at the same time point in a server folder; Step S752, traverse the data in each folder to perform a test, and perform the test in the order of view switching, BSD and DOW until the data in all folders are tested; Step S753, perform the next loop traversal and test until the shortcut exits.

Images