Motor vehicle headlamp light intensity detection method and device

Abstract

The invention discloses a motor vehicle headlamp light intensity detection method and device, and the method comprises the steps: obtaining a vehicle front edge structure of a target motor vehicle, and carrying out the matching of a first millimeter wave radar array of which the detection range covers the vehicle front edge of the target motor vehicle; the position of a target motor vehicle is obtained in real time through a first millimeter wave radar array, when it is judged that the position of the target motor vehicle reaches a first preset detection position, high beam light intensity detection is triggered, light intensity distribution of a headlamp of the target motor vehicle on a light detector screen is collected, and light spot mass center coordinates are calculated; whether the deviation of the light spot centroid coordinates relative to the corresponding expected light spot centroid coordinates exceeds a preset deviation amount or not is judged; if so, repositioning to obtain a second detection position; and continuously judging whether the deviation of the light spot centroid coordinate relative to the corresponding expected light spot centroid coordinate exceeds a preset deviation amount, if so, generating a headlamp light abnormity prompt, and otherwise, finishing light intensity detection. According to the invention, the accuracy of light intensity detection can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of motor vehicle headlamp detection, and in particular to a method and device for detecting the light intensity of motor vehicle headlamp light. Background Art

[0002] Motor vehicle headlight intensity testing is a crucial component in ensuring vehicle safety. With the increasing number of vehicles on the road, the development of headlight intensity testing technology is crucial for improving road safety and protecting drivers and pedestrians. Currently, this technology primarily adheres to the requirements of GB38900-2020, "Motor Vehicle Safety Technical Inspection Items and Methods." This technology uses photoelectric switches to block and locate the most prominent front edge of the vehicle, alerting inspectors when the vehicle has reached a designated stop, thereby completing the headlight intensity test.

[0003] Existing technologies typically employ various methods to address vehicle positioning, including through-beam or reflective photoelectric switches, ultrasonic sensors, and infrared ranging. Through-beam photoelectric switches determine vehicle position by detecting whether a beam is blocked; reflective photoelectric switches use changes in reflected light intensity for positioning; ultrasonic sensors calculate distance by transmitting and receiving ultrasonic signals; and infrared ranging devices use the principle of infrared reflection to determine the distance between the vehicle and the detection device. These methods meet detection requirements to a certain extent.

[0004] However, these conventional methods have significant drawbacks. Factors such as the reaction speed of the photoelectric switch and the response time of the inspector result in significant distance errors, making it impossible to precisely control the standard distance between the vehicle and the light meter's detection screen, thus affecting the accuracy of light intensity detection. Furthermore, manual re-adjustment of the detection distance is often required, which is time-consuming and tedious. This problem urgently requires an effective solution to improve detection accuracy and reduce the complexity of the positioning detection process. Summary of the Invention

[0005] In order to accurately locate and improve the accuracy of motor vehicle headlight intensity detection while reducing the complexity of the positioning detection process, the present application provides a motor vehicle headlight intensity detection method and device.

[0006] In a first aspect, the present application provides a method for detecting the light intensity of a motor vehicle headlamp, comprising: Real-time acquisition of images of target motor vehicles entering the detection area, and use of image analysis technology to obtain the vehicle front edge structure of the target motor vehicle; A first millimeter-wave radar array having a detection range covering the front edge of the target motor vehicle is matched according to the front edge structure of the target motor vehicle; the first millimeter-wave radar is mounted on the screen of the motor vehicle light detector; The position of the target motor vehicle is obtained by real-time collaborative positioning calculation using the matched first millimeter-wave radar array; when it is determined that the position of the target motor vehicle has reached the first preset detection position, a parking instruction for the target motor vehicle is generated; when the motor vehicle is parked, the high beam intensity detection is triggered, the light intensity distribution of the target motor vehicle headlight on the light detector screen is collected and the light spot centroid coordinates are calculated, and it is determined whether the calculated light spot centroid coordinates are offset by more than the preset offset compared with the expected light spot centroid coordinates corresponding to the first preset detection position; if so, the matched first millimeter-wave radar array is used for repositioning, the position data is obtained and recorded as the second detection position; otherwise, the light intensity distribution of the target motor vehicle headlight on the light detector screen is analyzed using image analysis technology to complete the light intensity detection; it is continued to be determined whether the calculated light spot centroid coordinates are offset by more than the expected light spot centroid coordinates corresponding to the second preset detection position; if so, a headlight abnormality prompt is generated; otherwise, the light intensity distribution of the target motor vehicle headlight on the light detector screen is analyzed using image analysis technology to complete the light intensity detection.

[0007] By adopting the above scheme, the differences in the leading edge structures of the detection vehicles are taken into consideration, and the first millimeter-wave radar array that can cover the leading edge of the target motor vehicle is matched to avoid detection errors caused by differences in the leading edges of the target motor vehicle; the coordinates of the center of mass of the light spot are calculated and compared with the expected center of mass coordinates of the light spot to accurately determine whether the target motor vehicle has arrived at the preset detection position, and re-position it, avoiding the positioning error problem caused by the reaction speed of the photoelectric switch and human operation in the traditional method; by re-positioning the expected center of mass coordinates of the light spot corresponding to the position and comparing the results with the actually calculated center of mass coordinates of the light spot, light intensity detection is directly completed without the user having to move the target vehicle again, effectively improving detection accuracy and efficiency.

[0008] Preferably, it also includes: When the position of the target vehicle is acquired in real time using the matched first millimeter-wave radar array, the offset angle of the target vehicle is calculated based on the distance difference between the light detector screen and the two front ends of the target vehicle; After calculating the light spot centroid coordinates and before comparing the calculated light spot centroid coordinates with the corresponding expected light spot centroid coordinates, the most recently acquired target motor vehicle position, target motor vehicle offset angle, and target motor vehicle front edge structure are input into a pre-built light spot centroid compensation network to obtain an output light spot centroid coordinate compensation value; the light spot centroid compensation network is constructed based on a deep learning algorithm and is trained and generated using historical target motor vehicle positions, historical target motor vehicle offset angles, historical target motor vehicle front edge structures, and corresponding historical light spot centroid coordinate compensation values; The compensated light spot centroid coordinates are calculated based on the acquired light spot centroid coordinate compensation values and used as the final calculated light spot centroid coordinates.

[0009] By adopting the above solution, taking into account that the vehicle will deviate when entering the detection area, combined with the light spot centroid compensation network, the compensated light spot centroid position is obtained according to the deviation angle. Without the need for vehicle adjustment, the vehicle positioning can be completed more accurately, avoiding detection errors caused by vehicle deviation, thereby improving the accuracy and reliability of light intensity detection.

[0010] Preferably, it also includes: Using image analysis technology to predetermine the preset detection lane where the target motor vehicle is located, and when using a matched first millimeter-wave radar array to obtain the target motor vehicle's position data in real time, using DBF digital beamforming technology to adjust the beam of the matched first millimeter-wave radar array so that the beam width of the matched first millimeter-wave radar array only covers the preset detection lane where the target motor vehicle is located; when using the adjusted first millimeter-wave radar array to transmit a beam to the target motor vehicle in the preset detection lane where the target motor vehicle is located, the transmitted beam is uniquely encoded, and when a reflected signal is received, the position data of the target motor vehicle in the preset detection lane is determined by decoding; While using the matched first millimeter-wave radar array to obtain the offset angle of the target motor vehicle in real time, which is recorded as the first offset angle, a second millimeter-wave radar array installed in the detection area and located on both sides of the actual target motor vehicle is used to calculate and obtain the offset angle based on the measured real-time distance, which is recorded as the second offset angle; the offset angle of the target motor vehicle is obtained by weighted calculation using the first offset angle and the second offset angle.

[0011] By adopting the above solution, DBF digital beamforming technology is used to adjust the beam width of the first millimeter-wave radar array to avoid interference signals from other detection lanes in the presence of parallel detection. The transmitted beam is uniquely encoded and decoded to determine the position data of the target vehicle when the reflected signal is received. This enables accurate differentiation of trigger signals during simultaneous multi-lane detection and improves detection accuracy. A second set of millimeter-wave radar arrays is used to measure the second offset angle of the target vehicle. Combined with the first offset angle, the offset angle measurement can be completed more accurately.

[0012] Preferably, it also includes: During repositioning, centroid coordinate pixel compensation processing and positioning anti-interference and optimization processing are performed; the positioning anti-interference and optimization processing includes: using environmental sensors to obtain environmental data of the detection area; using a deep learning algorithm to build an operating parameter acquisition model to obtain the first millimeter-wave radar operating parameters that best match the current environmental data; the operating parameter acquisition model is trained by using the first millimeter-wave radar operating parameters used when the historical positioning accuracy is greater than the preset accuracy under different environmental data conditions; and adjusting the operating parameters of the matched first millimeter-wave radar array according to the matched first millimeter-wave radar operating parameters.

[0013] By adopting the above solution, taking into account the influence of environmental factors on the collected light intensity distribution, repositioning is performed on the basis of completing the centroid coordinate pixel compensation and positioning anti-interference optimization processing to ensure the accuracy of the detection results.

[0014] Preferably, it also includes: Using a sensor installed on the target motor vehicle to obtain a vibration signal of the target motor vehicle in real time; A displacement model is constructed using a deep learning algorithm to obtain the vibration displacement distance of the leading edge of the target motor vehicle under the current vibration signal condition. The vibration offset distance of the leading edge of the target motor vehicle is used to correct the first millimeter-wave radar array position data to obtain the corrected first millimeter-wave radar array position data; and whether the target motor vehicle has arrived at the first preset detection position is determined based on the corrected first millimeter-wave radar array position data.

[0015] By adopting the above scheme, the influence of vibration interference of the target motor vehicle during the detection process on positioning is considered, the influence of vibration on the vehicle position is quantified, and the positioning is determined comprehensively, thereby improving positioning accuracy and detection accuracy.

[0016] Preferably, collecting the light intensity distribution of the target motor vehicle headlight on the light detector screen and calculating the coordinates of the centroid of the light spot includes: Preprocessing for light intensity distribution; Combined with deep learning technology, the light spot type is determined based on the pre-processed light intensity distribution. The light spot types include regular light spots, irregular light spots, and extremely irregular tube plates. If the result is a regular light spot, the weighted centroid method is used to calculate the coordinates of the center of mass of the light spot; If the judgment result is an irregular light spot, the coordinates of the center of mass of the light spot are calculated by using a regional processing method or an edge correction method; wherein, the regional processing method includes: using local maximum detection to identify multiple peak areas in the light spot, selecting the main peak based on the light intensity area or intensity ratio, and recalculating the center of mass within the main peak area; the edge correction method includes: using Canny edge detection to locate the boundary of the light spot, fitting the outline into a polygon, and calculating the geometric center as the coordinates of the center of mass of the light spot; if the judgment result is an extremely irregular light spot, the DBSCAN clustering algorithm is used to divide the high-density light intensity area, independently calculate the center of mass for each cluster, and then synthesize the global center of mass according to the light intensity weight as the coordinates of the center of mass of the light spot.

[0017] By adopting the above scheme, the collected light intensity distribution is classified and the light spot centroid coordinates are calculated based on different calculation methods to obtain more accurate light spot centroid coordinates to assist in subsequent precise positioning.

[0018] Preferably, obtaining the target motor vehicle position by collaborative positioning calculation includes: Count the number of first millimeter-wave radar arrays; Determining whether the number of first millimeter-wave radar arrays is greater than a first preset number; If it is greater than, calculating a first average value of the distances measured by each first millimeter-wave radar; retaining a first preset number of first millimeter-wave radar-measured distances that are sorted in ascending order of difference from the first average value of the distances; calculating a second average value of the retained first millimeter-wave radar-measured distances, and determining whether the largest difference from the second average value of the distances is less than a preset difference; if so, directly using the second average value as the measured distance; Otherwise, the distances measured by the first millimeter-wave radars of the second preset number that are ranked first are retained in ascending order of the difference from the second average value of the distances, and the actual position of the target motor vehicle is calculated using the triangulation algorithm; the second preset number is less than the first preset number and is the minimum number preset for the first millimeter-wave radar array.

[0019] By adopting the above scheme, according to the number of matched millimeter-wave radar arrays, the collaborative positioning algorithm is adaptively used to complete the calculation of the actual position of the target motor vehicle, effectively improving the accuracy of the target motor vehicle positioning.

[0020] In a second aspect, a device for detecting the light intensity of a motor vehicle headlamp includes: The target vehicle structure acquisition module is used to collect images of target vehicles entering the detection area in real time and obtain the vehicle front edge structure of the target vehicle using image analysis technology; a target vehicle positioning device matching module, configured to match a first millimeter-wave radar array having a detection range covering the front edge of the target vehicle according to the front edge structure of the target vehicle; the first millimeter-wave radar array being mounted on a screen of a vehicle light detector; The target motor vehicle collaborative positioning module is used to use the matched first millimeter wave radar array to coordinate positioning calculation in real time to obtain the target motor vehicle position; when it is determined that the position of the target motor vehicle has reached the first preset detection position, a target motor vehicle parking instruction is generated; the target motor vehicle positioning determination and light detection module is used to trigger high beam light intensity detection when the motor vehicle is parked, collect the light intensity distribution of the target motor vehicle headlight on the light detector screen and calculate the light spot centroid coordinates, and determine whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the first preset detection position; if so, reposition using the matched first millimeter wave radar array, obtain position data and record it as the second detection position; otherwise, use image analysis technology to analyze the light intensity distribution of the target motor vehicle headlight on the light detector screen to complete the light intensity detection; continue to determine whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the second preset detection position. If so, a headlight abnormality prompt is generated; otherwise, use image analysis technology to analyze the light intensity distribution of the target motor vehicle headlight on the light detector screen to complete the light intensity detection.

[0021] By adopting the above solution, the adaptive first millimeter-wave radar array based on the matching of the vehicle's leading edge structure can accurately obtain the position data of the target motor vehicle; the centroid coordinates of the collected light spot are calculated and compared with the expected centroid coordinates of the light spot, and repositioned according to the comparison result to obtain the repositioned corresponding expected centroid coordinates of the light spot. Without the user having to adjust the vehicle position, light intensity detection can be performed directly, thereby improving detection accuracy and automation, and ensuring the accuracy and reliability of light intensity detection results.

[0022] In a third aspect, the present application provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method as described above.

[0023] In a fourth aspect, the present application provides a computer device, comprising a memory, a processor, and a program stored and executable on the memory, wherein the program implements the steps of the above method when executed by the processor.

[0024] In summary, this application has the following beneficial effects: 1. The front edge structure of the target motor vehicle is acquired through image analysis technology, and the appropriate millimeter-wave radar array is matched to achieve adaptive position measurement for different vehicle models. This accurately determines whether the vehicle has reached the preset detection position, ensuring that the vehicle completes parking at the standard distance and triggers high-beam light intensity detection. The coordinates of the center of mass of the light spot are calculated and compared with the expected center of mass coordinates of the light spot, and repositioned to avoid errors caused by untimely parking operations by the user. Based on the repositioned position, the corresponding expected center of mass coordinates of the light spot are directly matched to complete light intensity detection directly, without the user having to readjust the parking position. This improves the accuracy of light intensity detection while reducing the complexity of the operation process. 2. Considering the offset of the target motor vehicle entering the detection target area, the spot centroid compensation network is used to compensate the target motor vehicle with the deviation angle, and then the compensated spot centroid coordinates are used to determine the positioning, effectively reducing the detection error caused by vehicle offset; 3. For the parallel detection of multiple target vehicles in multiple detection lanes, the millimeter-wave radar is used to adjust the beam and encode the beam to avoid interference from other detection lanes outside the preset detection lane where the target vehicle is located, and accurately measure the position of the target vehicle; multiple radar arrays are used to obtain multiple sets of offset angles in real time, and more accurate offset angles are obtained comprehensively to assist in the subsequent precise calculation of the center of mass coordinates of the light spot, thereby improving the accuracy of light intensity detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flow chart of a method for detecting the light intensity of a motor vehicle headlamp according to a specific embodiment; Figure 2 Schematic diagram of radar matrix detection in a method for detecting the light intensity of a motor vehicle headlamp according to a specific embodiment; Figure 3 Schematic diagram of parallel detection of multiple target vehicles in the method for detecting the light intensity of motor vehicle headlights described in a specific embodiment; Figure 4 Schematic diagram of a detection module in a motor vehicle headlamp light intensity detection device in a specific embodiment. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0027] like Figure 1 As shown, the embodiment of the present application discloses a method for detecting the light intensity of motor vehicle headlights, and the specific steps include: S1, collecting an image of a target motor vehicle entering a detection area, and analyzing and obtaining the vehicle front edge structure of the target motor vehicle.

[0028] Specifically, such as Figure 2 As shown, the target motor vehicle entering the detection area moves along the detection track toward the motor vehicle light detector screen, and the camera device (such as a CCD camera or a CMOS camera) installed on the motor vehicle light detector screen is used to obtain the image of the target motor vehicle entering the detection area in real time, and the image analysis technology is used to determine the preset detection lane where the target motor vehicle is located, such as lane 01 or lane 03 in the detection area, and the image analysis technology is used to analyze and obtain the vehicle leading edge structure of the target motor vehicle; wherein, the vehicle leading edge structure includes: a standard leading edge structure (ordinary cars and small trucks, etc.), a high leading edge structure (SUVs, off-road vehicles, etc.), a low leading edge structure (sports cars and sports cars, etc.) and a special leading edge structure (designed for vehicles with specific purposes, such as buses, trucks, etc.).

[0029] S2. Matching a first millimeter-wave radar array whose detection range covers the front edge of the target motor vehicle according to the front edge structure of the target motor vehicle.

[0030] Specifically, the first millimeter-wave radar is installed on the screen of the motor vehicle light detector and is distributed in layers, such as: lower radar, middle radar, upper radar, etc.

[0031] Pre-set matching rules based on experience match typical vehicle leading edge structures with photoelectric switches and radar arrays. Alternatively, a deep learning model can be constructed, using historical vehicle leading edge structures and a first millimeter-wave radar array that can cover the leading edge of these historical vehicles (quantifying the radar's detection range and the height of the vehicle's leading edge) as training data. This is then fed into the current target vehicle's leading edge structure to obtain a first millimeter-wave radar array that matches the target vehicle. For example, if the target vehicle is an SUV, which has a high leading edge structure, the first millimeter-wave radar array containing upper and middle radars will be matched accordingly.

[0032] This embodiment utilizes a 24GHz FMCW radar module, with four first-level millimeter-wave radars embedded in the four corners of the light meter's detection screen. This creates a multi-point ranging array, enabling a detection range of 0.2m-5m, with coverage of 1m±0.2m. Its range resolution is ≤3cm, supporting multi-target recognition and filtering out interference from metal reflections from vehicle bodies. Furthermore, to better cover varying vehicle front edge heights, the first-level millimeter-wave radar probe is tilted 15° and treated with an anti-reflective coating to minimize optical interference.

[0033] S3. Using the matched first millimeter-wave radar array, measure and determine whether the target motor vehicle has reached a first preset detection position, and park the vehicle and trigger light intensity detection based on the determination result.

[0034] Specifically, according to the relevant provisions of GB38900-2020 "Motor Vehicle Safety Technical Inspection Items and Methods", when manufacturers conduct high-beam headlight intensity testing on motor vehicles, they usually set the distance between the vehicle headlights and the light instrument detection screen to 1 meter. Therefore, the first preset detection position can be set to 1 meter, that is, when the measured target motor vehicle is 1 meter away from the light instrument detection screen, a parking instruction should be generated and the high-beam intensity detection should be triggered; among them, since the headlight intensity detection of motor vehicles is only for high-beam intensity detection, triggering the high-beam intensity detection refers to triggering the high-beam intensity detection.

[0035] Thus, the matching first millimeter-wave radar array transmits radar waves and receives reflected waves in real time, measuring and calculating the target vehicle's position data. The position data of the target vehicle measured and obtained by each first millimeter-wave radar in the first millimeter-wave radar array is collaboratively calculated to obtain the target vehicle's position. The collaborative positioning calculation can be performed using an average calculation method or a triangulation positioning algorithm.

[0036] When it is determined that the position of the target motor vehicle reaches the first preset detection position, a parking instruction for the target motor vehicle is generated; when it is determined that the motor vehicle completes the parking instruction, a high beam intensity detection is triggered.

[0037] In addition, considering that the number of first millimeter-wave radar arrays matching the front edge of the target vehicle is different for different target vehicles, in order to adaptively complete collaborative positioning, the target vehicle position is obtained by collaborative positioning calculation, which specifically includes: Count the number of first millimeter-wave radar arrays; in order to more accurately obtain the target motor vehicle position, at least a second preset number of millimeter-wave radar arrays are pre-set when matching the first millimeter-wave radar array, such as: the second preset number is 3.

[0038] Determine whether the number of the first millimeter-wave radar array is greater than a first preset number; the first preset number can be set according to user needs, such as: taking a value of 4 or greater than 4; the second preset number is less than the first preset number and is the minimum preset number of the first millimeter-wave radar array.

[0039] If it is greater than, a first average of the distances measured by each first millimeter-wave radar is calculated; a first preset number of first millimeter-wave radar distances are retained in ascending order of the difference between the distances measured by each first millimeter-wave radar and the first average of the distances; a second average of the retained first millimeter-wave radar distances is calculated, and it is determined whether the largest difference between the retained first millimeter-wave radar distances and the second average of the distances is less than a preset difference. If it is less than, the second average of the retained first millimeter-wave radar distances is directly used as the measured distance, thereby obtaining the target vehicle's position. For example, if the number of matched first millimeter-wave radar arrays is 4, and the four detection measurement data are stable within a 100ms window of 0.95-1.05m, all the first millimeter-wave radar distances D1, D2, D3, and D4 are directly retained, and a second average D is calculated. If the difference between D3 and D is the largest and greater than the preset difference, the distance between the target vehicle and the vehicle light detector screen is determined to be D.

[0040] Otherwise, that is, the number of first millimeter-wave radar arrays is less than 4 or the maximum difference between the measured distance of each retained first millimeter-wave radar and the second average value of the distance is not less than the preset difference, the distances measured by the first millimeter-wave radars with the second preset number of the first ones ranked first are retained in ascending order of the difference between the measured distance of each retained first millimeter-wave radar and the second average value of the distance, and the actual position of the target motor vehicle is calculated using the triangulation algorithm.

[0041] S4. Calculate the coordinates of the centroid of the light spot and compare them with the expected centroid coordinates of the light spot corresponding to the first preset detection position, obtain the offset and determine whether it exceeds the preset offset, and complete repositioning according to the determination result.

[0042] Specifically, when the high beam intensity detection is triggered, the target motor vehicle turns on the high beam headlights, collects the light intensity distribution of the target motor vehicle's headlights on the light detector screen, and calculates the coordinates of the centroid of the light spot. The specific calculation includes: Preprocessing is performed on the light intensity distribution; the preprocessing includes: using Gaussian filtering or median filtering to smooth the spot image, setting a dynamic threshold for the light intensity data (such as 10% to 20% of the maximum light intensity), filtering low signal-to-noise ratio areas, etc.

[0043] Combined with deep learning technology, the spot type is determined and acquired based on the pre-processed light intensity distribution. The spot types include regular spots, irregular spots, and extremely irregular tube plates. For example, a spot type acquisition neural network is constructed, with the model input being the light intensity distribution and the output being the spot type. The model is generated through training with historical light intensity distribution data and the corresponding labeled spot types.

[0044] If the result is a regular light spot, the weighted centroid method is used to calculate the centroid coordinates of the light spot; where the centroid coordinates (x c ,y c ) includes: Among them, I i is the pixel intensity point, (x i ,y i ) are pixel coordinates.

[0045] If the result of the determination is an irregular light spot, the coordinates of the light spot centroid are calculated by using a regional processing method or an edge correction method. The regional processing method includes: using local maximum detection or the Hessian matrix method to identify multiple peak areas in the light spot, selecting the main peak based on the light intensity area or intensity ratio, and recalculating the centroid within the main peak area. The edge correction method includes: using Canny edge detection or the light intensity gradient method to locate the light spot boundary, fitting the outline into a polygon, and calculating the geometric center as the coordinates of the light spot centroid. For extremely irregular light spots, the DBSCAN clustering algorithm is used to divide the high-density light intensity area, and the centroid of each cluster is calculated independently. The global centroid is then synthesized according to the light intensity weighting as the centroid coordinate of the light spot.

[0046] Compare the calculated coordinates of the center of mass of the light spot with the expected coordinates of the center of mass of the light spot corresponding to the first preset detection position, obtain the offset and compare the offset with the preset offset; determine whether it exceeds the preset offset, If it exceeds, it indicates that there may be a positioning error due to the user's parking operation, such as: the actual parking position is 0.96m away from the light tester screen, and the error is 0.04m; wherein, the expected light spot centroid coordinates corresponding to different preset detection positions are obtained by historically determining the light spot centroid coordinates projected onto the light tester screen by various types of vehicles that have passed the headlight intensity test at the corresponding specific preset detection position (such as 1 meter) and turning on the headlight high beam in the preset detection lane, wherein the relative position of the preset detection lane and the light tester screen is fixed; the headlight intensity test passing means that at a specific distance, the relationship between the light intensity received on the light tester screen and the minimum required luminous intensity of the headlight high beam at least satisfies the following formula: Where E is the received light intensity on the light detector screen, I is the minimum required high-beam headlight intensity, and d is the distance between the target vehicle and the light detector screen. The minimum required high-beam headlight intensity is set for each of the one-, two-, and four-lamp systems. Furthermore, the center of mass coordinates of the light spot are calculated separately for the left and right high-beam headlights of the target vehicle. If the offset between the calculated center of mass coordinates of the light spot and the expected center of mass coordinates corresponding to the first preset detection position exceeds the preset offset, an error is determined.

[0047] To avoid errors in light detection caused by positioning errors, the matching first millimeter-wave radar array is used for repositioning, and the position data is obtained and recorded as the second detection position, such as repositioning to a distance of 0.96m. Otherwise, it is determined that no error exists at present, and subsequent light intensity detection can be carried out directly, such as using a camera installed in the detection area to capture the light intensity distribution image of the target motor vehicle headlight on the light detector screen, and using image analysis technology to analyze the light intensity distribution of the target motor vehicle headlight on the light detector screen to complete the light intensity detection, including determining whether the light intensity meets the requirements; a spectrum analyzer can also be used to analyze the color temperature / lumen value.

[0048] S5. Continue to determine whether the calculated light spot centroid coordinates are offset from the expected light spot centroid coordinates corresponding to the second preset detection position by more than a preset offset, and complete the light intensity detection based on the determination result.

[0049] Specifically, continue to determine whether the calculated light spot centroid coordinates are offset from the expected light spot centroid coordinates corresponding to the second preset detection position, and compare whether the offset exceeds the preset offset; if it exceeds, it indicates that even if the positioning is re-determined, there is still a large light spot centroid coordinate error, and a corresponding headlamp abnormality prompt is generated; otherwise, it indicates that there is no large error, and image analysis technology is used to analyze the light intensity distribution of the target motor vehicle headlights on the light tester screen to complete the light intensity detection.

[0050] A specific embodiment differs from the above embodiment in that: an offset angle calculation function is added. Considering that a target vehicle entering the detection area will be offset, after the position and offset angle of the target vehicle are obtained by the millimeter wave radar array, the coordinates of the centroid of the light spot need to be compensated, and an offset angle correction prompt is generated in a timely manner. The method further includes: When the position of the target motor vehicle is acquired in real time using the matched first millimeter-wave radar array, the offset angle of the target motor vehicle is calculated synchronously based on the measured distance difference between the light detector screen and the two front ends of the target motor vehicle; wherein, the offset angle θ1 (denoted as the first offset angle) of the target motor vehicle is calculated based on the measured distance difference ΔL1 between the light detector screen and the two front ends of the target motor vehicle, and the formula is: tanθ1 = ΔX / |ΔL1|, where ΔX is the lateral width of the target vehicle.

[0051] After the light spot centroid coordinates are calculated and before the calculated light spot centroid coordinates are compared with the corresponding expected light spot centroid coordinates, the light spot centroid coordinates are compensated. In order to improve the compensation accuracy, the light spot centroid compensation network is selected for compensation. The light spot centroid compensation network is constructed based on a deep learning algorithm and is trained and generated through the historical target motor vehicle position, the historical target motor vehicle offset angle, the historical target motor vehicle vehicle front edge structure and the corresponding historical light spot centroid coordinate compensation value. The newly acquired target motor vehicle position, the target motor vehicle offset angle and the target motor vehicle vehicle front edge structure are input into the pre-constructed light spot centroid compensation network to obtain the output light spot centroid coordinate compensation value x e ,x e = L1·tanθ1; L1 is the distance from the light detector screen to the front end of the target vehicle; The coordinates of the center of mass of the light spot after compensation are calculated based on the obtained coordinates of the center of mass of the light spot, and the coordinates of the center of mass of the light spot after compensation are used as the final coordinates of the center of mass of the light spot (x c ,y c ); Specifically, since the offset of the target motor vehicle is usually in the horizontal direction and does not involve the vertical direction, the corresponding spot centroid has a lateral offset, and the calculated spot centroid coordinates are subtracted from the spot centroid coordinate compensation value x ec =x c -x e , get the coordinates of the center of mass of the light spot after compensation (x ec ,y ec ).

[0052] A specific embodiment differs from the above embodiment in that: beam coding technology is introduced to solve the interference problem during simultaneous multi-lane detection; specifically, the method further includes: like Figure 3As shown, to address the interference that can occur when using millimeter-wave radar to detect multiple target vehicles in parallel, image analysis technology is used to predetermine the target vehicle's location in a preset detection lane. To prevent positioning errors caused by the radar beam transmitted by the matched first millimeter-wave radar array covering not only the front end of the target vehicle but also target vehicles in adjacent lanes, DBF digital beamforming technology is used when acquiring the target vehicle's location data in real time using the matched first millimeter-wave radar array to adjust the matched first millimeter-wave radar array's beam width so that the matched first millimeter-wave radar array's beam width only covers the preset detection lane where the target vehicle is located. Furthermore, after receiving a reflected signal, the specific target vehicle is determined. When the adjusted first millimeter-wave radar array transmits a beam to a target vehicle in the preset detection lane where the target vehicle is located, the transmitted beam is uniquely encoded. Upon receiving the reflected signal, the location data of the target vehicle in the preset detection lane is determined through decoding. The preset detection lane ID of each target vehicle can be converted into a unique code, and decoding is then used to determine the specific target vehicle in the preset detection lane to which the received reflected signal belongs.

[0053] In addition to optimizing the measured distance, the measured offset angle is further optimized. When the offset angle of the target motor vehicle is obtained in real time using the matching first millimeter-wave radar array, which is recorded as the first offset angle, the second millimeter-wave radar array installed in the detection area and located on both sides of the actual target motor vehicle is used to calculate and obtain the offset angle based on the measured real-time distance, which is recorded as the second offset angle θ2; the second offset angle of the target motor vehicle is calculated using the real-time measured distance difference ΔL2, that is, sinθ2=ΔH / |ΔL2|, ΔH=v·Δt is the forward distance of the target motor vehicle, and v is obtained according to the measurement of the first millimeter-wave radar or through the vehicle bus; the offset angle of the target motor vehicle is obtained by weighted calculation using the first offset angle and the second offset angle, θ=αθ1+βθ2.

[0054] Furthermore, to address potential interference between different millimeter-wave radars, a master clock is established. Using time-division multiplexing, the transmission period of the first millimeter-wave radar array is interleaved with that of the second millimeter-wave radar array. For example, a master clock (with an accuracy of ±10ns) is established to interleave the transmission period of the first millimeter-wave radar (1ms) with the transmission period of the second millimeter-wave radar (0.5ms) to ensure there is no overlap. Furthermore, to further reduce interference, frequency band isolation optimization is implemented, using bandpass filters to achieve spectral isolation between the radars.

[0055] In a specific embodiment, considering the impact of environmental factors on the final light intensity detection result, the environmental factors are further considered during re-positioning to ensure the accuracy of light intensity detection. The system further includes: During repositioning, the center of mass coordinate pixel compensation processing, positioning anti-interference and optimization processing are performed, environmental data is collected and the radar operating parameters are adjusted to achieve more accurate positioning.

[0056] The pixel compensation process includes correcting the centroid coordinates according to the measured distance of the millimeter wave radar, and the formula is as follows: Where, X real is the centroid coordinate x C The value after compensation, d is the measurement distance of the millimeter wave radar, f is the focal length of the lens, Δ cal is the temperature compensation amount; The positioning anti-interference and optimization processing includes: taking into account the influence of environmental factors, using environmental sensors to obtain environmental data of the detection area; using a deep learning algorithm to build an operating parameter acquisition model to obtain the first millimeter-wave radar operating parameters that best match the current environmental data; the operating parameter acquisition model is trained by using the first millimeter-wave radar operating parameters used when the historical positioning accuracy is greater than the preset accuracy under different environmental data conditions; adjusting the operating parameters of the matched first millimeter-wave radar array according to the matched first millimeter-wave radar operating parameters; for example, when the humidity is greater than 80%, the optimal frequency band of 60GHZ is obtained, and the frequency band of the first millimeter-wave radar is adjusted to the optimal frequency band accordingly.

[0057] In addition, taking into account the possible existence of a radar failure, before the target motor vehicle arrives at the first preset detection position, the radar is pre-verified in combination with image recognition to determine whether a failure has occurred. An image of the target motor vehicle is collected, and when the front end of the target motor vehicle in the image reaches the third preset detection position or the fourth preset detection position, a determination is made as to whether the position obtained by measuring with the first millimeter-wave radar array corresponds to the third preset detection position or the fourth preset detection position. If so, it is determined that there is no failure in the current first millimeter-wave radar array. The third preset detection position and the fourth detection position are farther from the light detector screen than the first preset detection position.

[0058] In a specific embodiment, considering that the target motor vehicle may have positioning errors due to vibration, the method further includes: quantifying the impact of vibration on the vehicle position and effectively compensating for it to achieve more accurate positioning. Using a sensor installed on the target motor vehicle to obtain a vibration signal of the target motor vehicle in real time; A displacement offset model is constructed using a deep learning algorithm to obtain the vibration offset distance of the leading edge of the target motor vehicle under the vibration signal conditions at the current moment. The input of the model is the vibration signal, and the output is the vibration offset distance of the leading edge of the target motor vehicle. For example, if the input jitter is a high-frequency, low-amplitude vibration, such as a frequency range of 1-10Hz and an amplitude of ±5cm, the corresponding vibration offset distance is a lateral offset of ±0.02m.

[0059] The vibration offset distance of the leading edge of the target motor vehicle is used to correct the first millimeter-wave radar array position data to obtain the corrected first millimeter-wave radar array position data; and whether the target motor vehicle has arrived at the first preset detection position is determined based on the corrected first millimeter-wave radar array position data.

[0060] In addition, in order to better eliminate the influence of vibration, Kalman filter fusion technology can also be used to design an extended Kalman filter, and the state variables are the actual position (x, y) and speed (v of the output target vehicle x , v y ); Input observation values, including: the first millimeter wave radar array measurement distance d and azimuth angle θ (converted to Cartesian coordinates x r =dcosθ,y r =dsinθ), the trigger position x of the photoelectric switch p ; Update the state. The state update equation includes: Among them, w x With w y , which is used to describe the random disturbance caused by vibration. The corresponding disturbance value can be obtained by quantifying the vibration.

[0061] like Figure 4 As shown, the embodiment of the present application discloses a device for detecting the light intensity of a motor vehicle headlamp, specifically comprising: The target vehicle structure acquisition module 101 is used to collect images of target vehicles entering the detection area in real time and obtain the vehicle front edge structure of the target vehicle using image analysis technology; The target vehicle positioning device matching module 102 is configured to match a first millimeter-wave radar array having a detection range covering the front edge of the target vehicle based on the front edge structure of the target vehicle; the first millimeter-wave radar array is mounted on the screen of the vehicle light detector and is distributed in layers; The target vehicle collaborative positioning module 103 is configured to obtain the target vehicle position by using the matched first millimeter-wave radar array for real-time collaborative positioning calculation; and generate a parking instruction for the target vehicle when it is determined that the target vehicle position has reached the first preset detection position; The target motor vehicle positioning determination and light detection module 104 is used to trigger high beam light intensity detection when the motor vehicle is parked, collect the light intensity distribution of the target motor vehicle headlight on the light detector screen and calculate the light spot centroid coordinates, and determine whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the first preset detection position; if so, reposition using the matched first millimeter wave radar array, obtain position data and record it as the second detection position; otherwise, use image analysis technology to analyze the light intensity distribution of the target motor vehicle headlight on the light detector screen to complete the light intensity detection; continue to determine whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the second preset detection position; if so, generate a headlight abnormality prompt; otherwise, use image analysis technology to analyze the light intensity distribution of the target motor vehicle headlight on the light detector screen to complete the light intensity detection.

[0062] In a specific embodiment, the system further includes: The target vehicle offset angle acquisition module 105 is configured to simultaneously calculate the target vehicle offset angle based on the distance difference between the light detector screen and the two front ends of the target vehicle when acquiring the target vehicle position in real time using the matched first millimeter-wave radar array; The target motor vehicle positioning determination and light detection module 104 is further used to input the latest acquired target motor vehicle position, the target motor vehicle offset angle and the target motor vehicle front edge structure into a pre-built light spot centroid compensation network after calculating the light spot centroid coordinates and before comparing the calculated light spot centroid coordinates with the corresponding expected light spot centroid coordinates, so as to obtain an output light spot centroid coordinate compensation value; the light spot centroid compensation network is constructed based on a deep learning algorithm and is trained and generated through historical target motor vehicle positions, historical target motor vehicle offset angles, historical target motor vehicle front edge structures and corresponding historical light spot centroid coordinate compensation values; the compensated light spot centroid coordinates are calculated based on the acquired light spot centroid coordinate compensation values as the final calculated light spot centroid coordinates.

[0063] In a specific embodiment, the target motor vehicle collaborative positioning module 103 is further used to use image analysis technology to pre-determine the preset detection lane where the target motor vehicle is located. When using the matched first millimeter wave radar array to obtain the position data of the target motor vehicle in real time, the DBF digital beamforming technology is used to adjust the beam of the matched first millimeter wave radar array so that the beam width of the matched first millimeter wave radar array only covers the preset detection lane where the target motor vehicle is located; when using the adjusted first millimeter wave radar array to transmit a beam to the target motor vehicle in the preset detection lane where the target motor vehicle is located, the transmitted beam is uniquely encoded, and when the reflected signal is received, the position data of the target motor vehicle in the corresponding preset detection lane is determined by decoding.

[0064] The target motor vehicle offset angle acquisition module 105 is further configured to, while utilizing the matched first millimeter-wave radar array to acquire the offset angle of the target motor vehicle in real time, which is recorded as a first offset angle, utilize a second millimeter-wave radar array installed within the detection area and located on both sides of the actual target motor vehicle to calculate and acquire the offset angle based on the measured real-time distance, which is recorded as a second offset angle; and utilize a weighted calculation of the first offset angle and the second offset angle to acquire the offset angle of the target motor vehicle.

[0065] In a specific embodiment, the target motor vehicle positioning determination and light detection module 104 in the system is also used to perform centroid coordinate pixel compensation processing and positioning anti-interference and optimization processing during repositioning; the positioning anti-interference and optimization processing includes: using environmental sensors to obtain environmental data of the detection area; using a deep learning algorithm to build an operating parameter acquisition model to obtain the first millimeter-wave radar operating parameters that best match the current environmental data; the operating parameter acquisition model is trained by using the first millimeter-wave radar operating parameters used when the historical positioning accuracy is greater than the preset accuracy under different environmental data conditions; and adjusting the operating parameters of the matched first millimeter-wave radar array according to the matched first millimeter-wave radar operating parameters.

[0066] In a specific embodiment, the target motor vehicle positioning determination and light detection module 104 is further used to obtain the vibration signal of the target motor vehicle in real time using a sensor installed on the target motor vehicle; use a deep learning algorithm to construct a displacement offset model to obtain the vibration offset distance of the front edge of the target motor vehicle under the vibration signal condition at the current moment; use the vibration offset distance of the front edge of the target motor vehicle to correct the first millimeter wave radar array to obtain the corrected first millimeter wave radar array to obtain the position data; and determine whether the target motor vehicle has arrived at the first preset detection position based on the corrected first millimeter wave radar array to obtain the position data.

[0067] The embodiment of the present application also discloses a computer-readable storage medium.

[0068] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed by the above-mentioned motor vehicle headlight intensity detection method. The computer-readable storage medium includes, for example: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0069] The embodiment of the present application also discloses a computer device.

[0070] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded by the processor and execute the above-mentioned motor vehicle headlamp light intensity detection method.

[0071] The above are all preferred embodiments of the present application and are not intended to limit the scope of protection of this application. Unless otherwise stated, any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features. In other words, unless otherwise stated, each feature is merely an example of a series of equivalent or similar features.

Claims

1. A method for detecting the light intensity of a motor vehicle headlamp, characterized in that: include: Real-time acquisition of images of target motor vehicles entering the detection area, and use of image analysis technology to obtain the vehicle front edge structure of the target motor vehicle; A first millimeter-wave radar array having a detection range covering the front edge of the target motor vehicle is matched according to the front edge structure of the target motor vehicle; the first millimeter-wave radar is mounted on the screen of the motor vehicle light detector; The target motor vehicle position is obtained by using a matched first millimeter-wave radar array for real-time collaborative positioning calculation; when it is determined that the target motor vehicle position has reached a first preset detection position, a parking instruction for the target motor vehicle is generated; when the motor vehicle is parked, a high-beam light intensity detection is triggered, the light intensity distribution of the target motor vehicle headlights on the light detector screen is collected, and the coordinates of the light spot centroid are calculated, and it is determined whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the first preset detection position; If it exceeds, the matching first millimeter-wave radar array is used to reposition, obtain position data and record it as the second detection position; otherwise, the image analysis technology is used to analyze the light intensity distribution of the target motor vehicle headlights on the light detector screen to complete the light intensity detection; continue to judge whether the calculated light spot centroid coordinates are offset by more than the preset offset compared to the expected light spot centroid coordinates corresponding to the second preset detection position. If so, a headlight abnormality prompt is generated; otherwise, the image analysis technology is used to analyze the light intensity distribution of the target motor vehicle headlights on the light detector screen to complete the light intensity detection.

2. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: Also includes: When the position of the target vehicle is acquired in real time using the matched first millimeter-wave radar array, the offset angle of the target vehicle is calculated based on the distance difference between the light detector screen and the two front ends of the target vehicle; After calculating the light spot centroid coordinates and before comparing the calculated light spot centroid coordinates with the corresponding expected light spot centroid coordinates, the newly acquired target vehicle position, the target vehicle offset angle, and the vehicle leading edge structure of the target vehicle are input into a pre-built light spot centroid compensation network to obtain an output light spot centroid coordinate compensation value; The light spot centroid compensation network is constructed based on a deep learning algorithm and is trained and generated through the historical target vehicle position, the historical target vehicle offset angle, the historical target vehicle front edge structure, and the corresponding historical light spot centroid coordinate compensation value; The compensated light spot centroid coordinates are calculated based on the acquired light spot centroid coordinate compensation values and used as the final calculated light spot centroid coordinates.

3. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: Also includes: Using image analysis technology to predetermine the preset detection lane where the target motor vehicle is located, and when using a matched first millimeter-wave radar array to obtain the target motor vehicle's position data in real time, using DBF digital beamforming technology to adjust the beam of the matched first millimeter-wave radar array so that the beam width of the matched first millimeter-wave radar array only covers the preset detection lane where the target motor vehicle is located; when using the adjusted first millimeter-wave radar array to transmit a beam to the target motor vehicle in the preset detection lane where the target motor vehicle is located, the transmitted beam is uniquely encoded, and when a reflected signal is received, the position data of the target motor vehicle in the preset detection lane is determined by decoding; While using the matched first millimeter-wave radar array to obtain the offset angle of the target motor vehicle in real time, which is recorded as the first offset angle, a second millimeter-wave radar array installed in the detection area and located on both sides of the actual target motor vehicle is used to calculate and obtain the offset angle based on the measured real-time distance, which is recorded as the second offset angle; the offset angle of the target motor vehicle is obtained by weighted calculation using the first offset angle and the second offset angle.

4. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: Also includes: During repositioning, centroid coordinate pixel compensation processing, positioning anti-interference and optimization processing are performed; The positioning anti-interference and optimization processing includes: using environmental sensors to obtain environmental data of the detection area; using a deep learning algorithm to build an operating parameter acquisition model to obtain the first millimeter-wave radar operating parameters that best match the current environmental data; the operating parameter acquisition model is trained by using the first millimeter-wave radar operating parameters used when the historical positioning accuracy is greater than the preset accuracy under different environmental data conditions; and adjusting the operating parameters of the matched first millimeter-wave radar array according to the matched first millimeter-wave radar operating parameters.

5. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: Also includes: Using a sensor installed on the target motor vehicle to obtain a vibration signal of the target motor vehicle in real time; A displacement model is constructed using a deep learning algorithm to obtain the vibration displacement distance of the leading edge of the target motor vehicle under the current vibration signal condition. Correcting the position data acquired by the first millimeter-wave radar array using the vibration offset distance of the leading edge of the target motor vehicle, and obtaining the corrected position data acquired by the first millimeter-wave radar array; The position data obtained by the corrected first millimeter-wave radar array is used to determine whether the target motor vehicle has arrived at the first preset detection position.

6. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: The collecting of the light intensity distribution of the target motor vehicle headlight on the light detector screen and the calculation of the coordinates of the centroid of the light spot include: Preprocessing for light intensity distribution; Combined with deep learning technology, the light spot type is determined based on the pre-processed light intensity distribution. The light spot types include regular light spots, irregular light spots, and extremely irregular tube plates. If the result is a regular light spot, the weighted centroid method is used to calculate the coordinates of the center of mass of the light spot; If the result of the determination is an irregular light spot, the coordinates of the light spot centroid are calculated by using a regional processing method or an edge correction method. The regional processing method includes: using local maximum detection to identify multiple peak areas in the light spot, selecting the main peak based on the light intensity area or intensity ratio, and recalculating the centroid within the main peak area. The edge correction method includes: using Canny edge detection to locate the light spot boundary, fitting the outline into a polygon, and calculating the geometric center as the coordinates of the light spot centroid. For extremely irregular light spots, the DBSCAN clustering algorithm is used to divide the high-density light intensity area, and the centroid of each cluster is calculated independently. The global centroid is then synthesized according to the light intensity weighting as the centroid coordinate of the light spot.

7. The method for detecting the light intensity of a motor vehicle headlamp according to claim 1, wherein: The method of obtaining the target vehicle position by collaborative positioning calculation includes: Count the number of first millimeter-wave radar arrays; Determining whether the number of first millimeter-wave radar arrays is greater than a first preset number; If it is greater than, calculating a first average value of the distances measured by each first millimeter-wave radar; retaining a first preset number of first millimeter-wave radar-measured distances that are sorted in ascending order of difference from the first average value of the distances; calculating a second average value of the retained first millimeter-wave radar-measured distances, and determining whether the largest difference from the second average value of the distances is less than a preset difference; if so, directly using the second average value as the measured distance; Otherwise, the distances measured by the first millimeter-wave radars of the second preset number that are ranked first are retained in ascending order of the difference from the second average value of the distances, and the actual position of the target motor vehicle is calculated using the triangulation algorithm; the second preset number is less than the first preset number and is the minimum number preset for the first millimeter-wave radar array.

8. A device for detecting the light intensity of a motor vehicle headlight, characterized in that: include: The target vehicle structure acquisition module is used to collect images of target vehicles entering the detection area in real time and obtain the vehicle front edge structure of the target vehicle using image analysis technology; a target vehicle positioning device matching module, configured to match a first millimeter-wave radar array having a detection range covering the front edge of the target vehicle according to the front edge structure of the target vehicle; the first millimeter-wave radar array being mounted on a screen of a vehicle light detector; The target motor vehicle collaborative positioning module is used to obtain the target motor vehicle position by using the matched first millimeter wave radar array for real-time collaborative positioning calculation; when it is determined that the position of the target motor vehicle has reached the first preset detection position, a parking instruction for the target motor vehicle is generated; a target motor vehicle positioning determination and light detection module, configured to trigger high-beam light intensity detection when the motor vehicle is parked, collect the light intensity distribution of the target motor vehicle's headlights on the light detector screen, calculate the coordinates of the light spot centroid, and determine whether the calculated light spot centroid coordinates are offset by more than a preset offset compared to the expected light spot centroid coordinates corresponding to the first preset detection position; If it exceeds, the matching first millimeter-wave radar array is used to reposition, obtain position data and record it as the second detection position; otherwise, the image analysis technology is used to analyze the light intensity distribution of the target motor vehicle headlights on the light detector screen to complete the light intensity detection; continue to judge whether the calculated light spot centroid coordinates are offset by more than the preset offset compared to the expected light spot centroid coordinates corresponding to the second preset detection position. If so, a headlight abnormality prompt is generated; otherwise, the image analysis technology is used to analyze the light intensity distribution of the target motor vehicle headlights on the light detector screen to complete the light intensity detection.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer device, characterized in that: The computer device includes a memory, a processor, and a program stored and executable on the memory, and when the program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images