Vehicle driving environment sensing method and anti-glare control method of DLP intelligent vehicle lamp

Abstract

The present application belongs to the technical field of vehicle lighting, and particularly relates to a vehicle driving environment sensing method and a glare control method for a DLP intelligent vehicle lamp. The vehicle driving environment sensing method comprises: collecting road condition information in a shooting range by a binocular camera, and transmitting the information as a to-be-detected image to a target detection model deployed on a vehicle-mounted computer in real time; predicting the to-be-detected image by the target detection model to detect a set target and a pixel coordinate position thereof in the to-be-detected image; and calculating a distance from the set target in the to-be-detected image to the binocular camera by a ranging algorithm. The present application can accurately identify pedestrians and their positions by real-time collection and analysis of road condition information in front of a vehicle in combination with a target detection model, and can calculate a target distance by using a stereo vision technology, so as to provide accurate basis for light mode adjustment of a vehicle lamp by quickly determining a spatial relationship between a target object and the vehicle, thereby effectively avoiding the problem of dazzling pedestrians and oncoming vehicles.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of vehicle lighting, and particularly relates to a vehicle driving environment perception method and a glare control method of a DLP intelligent vehicle lamp. BACKGROUND

[0002] With the rapid development of automotive intelligent technology, the functions of vehicle-mounted systems gradually change from traditional passive assistance to active intelligence, especially in night driving environments, intelligent control of vehicle lamps has become one of the key research fields to improve driving safety and driving comfort. The traditional automotive lighting system mainly relies on fixed light intensity distribution design, which is difficult to dynamically adjust the lighting mode according to the actual road conditions and target objects such as pedestrians or oncoming vehicles. This fixed light intensity design often leads to the following problems in actual driving:

[0003] In night driving, when the vehicle light directly illuminates pedestrians, the strong light beam can cause temporary blindness or glare of the pedestrians, especially in urban streets or densely populated areas, which may cause visual discomfort to pedestrians, and even induce traffic accidents; high-intensity high beams can easily cause driver visual fatigue or temporary disability when oncoming vehicles enter the line of sight, which seriously affects night driving safety; the lighting mode of the current vehicle lamp system mostly relies on simple switching, such as switching between low beam and high beam, and lacks the ability to perceive real-time road conditions, especially in complex driving scenarios such as multiple pedestrians, curves, slopes, etc. The light cannot be precisely controlled.

[0004] In the prior art, some solutions based on monocular cameras or infrared sensors attempt to achieve local light shielding or light adjustment by identifying target areas, but these methods have limited perception range, insufficient accuracy, and poor environmental adaptability. For example, the performance of monocular cameras in low-light environments will decrease significantly, and infrared sensors have limited ability to recognize target shapes, which cannot provide sufficient optical shielding accuracy. At the same time, these solutions often have difficulty in effectively distinguishing different road objects such as pedestrians and obstacles, resulting in inflexible light control strategies, which further limits the actual application effect of the system. SUMMARY

[0005] The present application provides a vehicle driving environment perception method and a glare control method of a DLP intelligent vehicle lamp, which can effectively solve the problems in the background art.

[0006] To achieve the above-mentioned purposes, the technical solutions of the present application are as follows:

[0007] The vehicle driving environment perception method is realized based on a binocular camera, the binocular camera is installed at a set position of the vehicle body, and the method comprises:

[0008] The binocular camera collects road condition information in a shooting range, and the collection result is transmitted to a target detection model deployed on a vehicle-mounted computer as a to-be-detected image in real time;

[0009] The target detection model is used to predict the to-be-detected image to detect a set target and a pixel coordinate position of the set target in the to-be-detected image;

[0010] A ranging algorithm is used to calculate the distance from the set target in the to-be-detected image to the binocular camera.

[0011] Further, the implementation steps of the ranging algorithm include:

[0012] For each pixel of the to-be-detected image, the value of the pixel matching cost is obtained through the fusion of the costs processed by the BT cost and the SobelX operator;

[0013] The value of each pixel is locally smoothed to obtain a cost matrix;

[0014] Multi-path constraint cost aggregation is performed on the cost matrix to obtain an aggregated cost matrix;

[0015] The optimal disparity value of each pixel is determined based on the aggregated cost matrix, and a disparity map is obtained through post-processing optimization;

[0016] Distance information is calculated based on the disparity map.

[0017] Further, the calculation steps of the BT cost include:

[0018] The gray value of the sub-pixel position (x R +x, y) between the pixels (x R -n, y) and (x R +n, y) in the input left eye image is calculated.

[0019] The gray value of the sub-pixel position (x R +d+x, y) between the pixels (x R +d-n, y) and (x R +d+n, y) in the input right eye image is calculated.

[0020] The BT cost is calculated by the following formula:

[0021]

[0022] cos = min (cos1, cos2);

[0023] wherein cos is the BT cost, cos1 and cos2 are bidirectional cost values, d is the disparity, 0.1n0.9, I R (x R , y) is a gray value at a sub-pixel position (x R , y) in a left eye image, IT(x R , d, y) is a gray value at a sub-pixel position (x R +d, y) offset by a disparity d in a right eye image.

[0024] The glare control method of the DLP intelligent vehicle lamp comprises the following steps:

[0025] The vehicle driving environment perception method is used to perceive the environment in front of the vehicle to obtain a perception result, which includes a set target and its pixel coordinate position in a to-be-detected image, and a distance from the set target to the binocular camera.

[0026] The correspondence between the set target and the light area of the vehicle lamp is established in combination with the optical characteristics of the DLP intelligent vehicle lamp, and a region requiring brightness adjustment is determined.

[0027] The DMD control chip is used to independently control a plurality of micromirrors to adjust the brightness of each region.

[0028] Further, in the process of determining the region requiring brightness adjustment, the environmental light intensity is combined to determine.

[0029] Further, the DMD control chip is used to independently control a plurality of micromirrors to adjust the brightness of each region, including:

[0030] According to the perception result and the environmental light intensity, the brightness requirement of each light area is calculated.

[0031] The brightness requirement of each light area is converted into a corresponding PWM signal to generate a duty cycle.

[0032] The PWM signal is used to drive the micromirror switch to adjust the light area brightness.

[0033] Further, the environmental light intensity is detected by a light sensor and / or identified through a to-be-detected image collected by the binocular camera.

[0034] Further, the environmental light intensity is identified through a to-be-detected image collected by the binocular camera, including:

[0035] The left and right images are collected by the binocular camera and time and space alignment is performed.

[0036] Obtaining a set target and its pixel coordinate position detected in the to-be-detected image by using the target detection model, and eliminating the set target from the left and right images;

[0037] Converting the remaining part of the left and right images after elimination into a gray image;

[0038] Calculating an average gray value as the current ambient light intensity.

[0039] Further, the set target is eliminated from the left and right images, specifically:

[0040] Replacing the pixel value of the region corresponding to the set target with a set value;

[0041] Or, marking the region corresponding to the set target as invalid and not participating in the brightness calculation.

[0042] Further, the ambient light intensity is detected by the light sensor and identified by the to-be-detected image collected by the binocular camera, including:

[0043] Based on the detection result of the light sensor, when the difference between the current and previous detection results is greater than a set value, the identification of the to-be-detected image is started;

[0044] According to the identification result of the to-be-detected image, the basis is revised to obtain the final ambient light intensity.

[0045] Through the technical solution of the present application, the following technical effects can be achieved:

[0046] In the present application, an environment perception method based on a binocular camera is proposed, and a glare control method of a DLP intelligent car lamp is proposed based on the method; by real-time acquisition and analysis of road condition information in the set direction of the vehicle, the position of the pedestrian is accurately identified in combination with a target detection model, and the target distance is calculated by using a stereo vision technology, which can provide data support for dynamic adjustment of the intelligent car lamp; through the parallax map-based ranging algorithm, the present application can quickly determine the spatial relationship between the target object and the vehicle, and provide accurate basis for light mode adjustment of the car lamp, thereby effectively avoiding the glare problem of the pedestrian and the oncoming vehicle, and the DLP intelligent car lamp can perform pixel-level accurate control on the illumination area through the control method in the present application, avoid the limitations of traditional mechanical shading, has fast response speed, and can respond to the change of the road condition in real time, while ensuring safe driving and providing the best lighting effect. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only represent some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0048] Figure 1 Flow chart for vehicle driving environment perception method;

[0049] Figure 2 Flow chart for ranging algorithm implementation;

[0050] Figure 3 Flow chart for glare control method of DLP intelligent car light;

[0051] Figure 4 Flow chart for adjusting the brightness of each region;

[0052] Figure 5 Flow chart for identifying the ambient light intensity through the image to be detected collected by the binocular camera. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments only represent some embodiments of the present application, not all embodiments.

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0055] Embodiment one:

[0056] As shown in the figure, the vehicle driving environment perception method is realized based on a binocular camera, the binocular camera is installed at a set position of the vehicle body, and the method comprises: Figure 1

[0057] S1: acquiring road condition information in the shooting range through the binocular camera, and transmitting the acquisition result as a to-be-detected image to a target detection model deployed on a vehicle-mounted computer in real time;

[0058] S2: predicting the to-be-detected image through the target detection model to detect a set target and its pixel coordinate position in the to-be-detected image;

[0059] ​S3: a ranging algorithm is used to calculate the distance from the set target in the image to be detected to the binocular camera.

[0060] In the implementation process, the binocular camera can be installed at the front position of the vehicle body, or at the rear or other positions of the vehicle body. Regardless of the installation position, the environmental perception of the corresponding direction can be realized, and thus is within the protection scope of the present application.

[0061] In the embodiment, an environmental perception method based on a binocular camera is proposed. In a specific implementation mode in which the binocular camera is installed at the front position of the vehicle, the road condition information in front of the vehicle is collected and analyzed in real time, and the target detection model is combined to accurately identify pedestrians and corresponding direction vehicles and their positions. The target distance is calculated by using the stereo vision technology, which can provide data support for the dynamic adjustment of the intelligent car light. Through the ranging algorithm based on the parallax map, the spatial relationship between the target object and the vehicle can be quickly determined, which provides accurate basis for the light mode adjustment of the car light, thereby effectively avoiding the dazzling problem of pedestrians and oncoming vehicles.

[0062] In the implementation process, the binocular camera simultaneously collects two synchronized images, i.e., a left image and a right image, which need to be calibrated by a time stamp to ensure that the two images are completely aligned in time. During image acquisition, it may be affected by environmental factors such as light intensity and weather conditions. Therefore, preprocessing can be performed during the acquisition process, such as: light equalization, using CLAHE to improve the visibility of the image; noise reduction processing, reducing image noise by Gaussian filtering or median filtering. The collected images can be in standard data formats such as JPEG, PNG or RAW, and transmitted to the vehicle-mounted computer through a high-speed bus or through a dedicated communication protocol to ensure real-time performance.

[0063] The target detection model used in the embodiment can be based on a deep learning algorithm, such as YOLO, SSD or Faster R-CNN. Selecting a suitable model requires balancing real-time performance and accuracy. As a specific implementation mode, the yolov9m lightweight target detection model can be specifically used. The yolov9m lightweight target detection model is trained based on the night traffic target data set in the embodiment, covering motor vehicles, non-motor vehicles, pedestrians and other targets. The model obtained by training predicts the real-time driving picture obtained by the vehicle-mounted binocular camera. Taking the installation of the binocular camera at the front position as an example, the motor vehicles, non-motor vehicles and pedestrian targets in front and their pixel coordinate positions can be detected.

[0064] For step S3, as a specific implementation mode, as shown in Figure 2 The implementation steps of the ranging algorithm include:

[0065] A1: for each pixel of the image to be detected, a pixel matching cost value is obtained by fusing a BT cost and a cost processed by a Sobel X operator; the BT cost is directly calculated from the input image, thereby retaining more original image information, and the cost processed by the Sobel X operator retains more edge and detail information; fusing the cost processed by the Sobel X operator with the cost directly calculated from the original image can improve the accuracy of the cost value;

[0066] For the cost fusion in this step, as a specific example, in the implementation process, the cost weight adjustment method can be used, for example, the BT cost is a first cost, the weight of which is W1, and the cost processed by the Sobel X operator is a second cost, the weight of which is W2; in the fusion process of the first cost and the second cost, the weight can be adjusted by experiment or self-adaption, and the fusion formula is:

[0067] C(p, d) = W1·C BT (p, d) + W2·C Sobel (p, d);

[0068] Wherein, C(p, d) is the fused cost value, used to represent the matching degree of pixel p at disparity d; C BT (p, d) represents the cost value calculated by using the BT method; C Sobel (p, d) represents the cost value processed by using the Sobel operator. Of course, this fusion method is only a specific example, and other fusion methods, such as maximum fusion, minimum fusion, product fusion and geometric mean fusion, etc., are also within the protection scope of the present application, as long as they can meet the technical purpose of the present application.

[0069] Before the implementation of this step, the image can be first subjected to light balance and denoising processing to improve the robustness of cost calculation;

[0070] A2: locally smoothing the cost value of each pixel to obtain a cost matrix;

[0071] As a specific smoothing method, the obtained cost value can be subjected to block processing, in which the cost value of each pixel can be replaced by the sum of the surrounding neighborhood cost values to improve the robustness of matching; and the neighborhood window size can be dynamically adjusted according to the texture complexity of the region where the pixel is located, such as 3x3, 5x5, etc., preferably using a large window in a sparse texture region and using a small window in a dense texture region;

[0072] A3: performing multi-path constraint cost aggregation on the cost matrix to obtain an aggregated cost matrix; this step can achieve the effect of global energy function minimization similar to a global stereo matching algorithm; in the implementation process, in order to balance the accuracy and real-time performance of the algorithm matching, four-way aggregation can be specifically performed on the cost matrix, i.e. horizontal, vertical and two diagonal path;

[0073] A4: determining the optimal disparity value of each pixel based on the aggregated cost matrix, and obtaining a disparity map through post-processing optimization; specifically, the disparity corresponding to the minimum cost value can be selected as the optimal disparity value from the cost values of all disparities of the pixel; disparity optimization includes confidence detection, sub-pixel disparity value and left-right consistency detection, etc.

[0074] A5: calculating distance information based on the disparity map, i.e. obtaining the distance information between the binocular camera and the set target.

[0075] As a preferred embodiment of the above, the calculation step of the BT cost includes:

[0076] calculating the gray value of the sub-pixel position (x R +x, y) between the pixels (x R -n, y) and (x R +n, y) in the input left eye image

[0077] calculating the gray value of the sub-pixel position (x R +d+x, y) between the pixels (x R +d-n, y) and (x R +d+n, y) in the input right eye image

[0078] The BT cost is calculated by the following formula:

[0079]

[0080] cos = min (cos1, cos2);

[0081] wherein, cos is the BT cost, cos1 and cos2 are bidirectional cost values, d is the disparity, 0.1 ≤ n ≤ 0.9, I R (x R , y) is the gray value of the sub-pixel position (x R , y) in the left eye image, I T (x R +d, y) is the gray value of the sub-pixel position (x R +d, y) in the right eye image offset by the disparity d.

[0082] In the preferred scheme, the sub-pixel gray value extraction based on left and right images is used to calculate the bidirectional matching cost values cos1 and cos2, which further improves the accuracy of pixel matching while preserving image details and global features, especially in sparse texture or repetitive texture scenes, which helps to alleviate the disparity calculation error and generate a higher quality disparity map. In the implementation process, by calculating the gray value difference of the target pixel in the left and right images, combining the bidirectional matching cost cos1 and cos2, and taking the minimum value of the cost as the final BT cost, the influence of noise and mismatch is effectively reduced. In the case of weak light environment, complex texture background or dynamic target, this scheme can more stably match the image features and ensure the reliability of the ranging algorithm. Even in dynamic scenes with large changes in illumination, the refined gray value interpolation calculation can enhance the algorithm's ability to capture target details, thereby reducing the problem of mismatch caused by changes in illumination. The disparity optimization strategy based on the optimal cost value selection in the scheme can complete the full process of cost calculation with low computational cost, which is suitable for the high real-time requirements of vehicle-mounted systems. At the same time, the calculation method of BT cost maintains high accuracy while having moderate computational complexity, which is conducive to real-time operation on embedded hardware platforms.

[0083] Embodiment Two

[0084] As shown in Figure 3 , the glare control method of the DLP intelligent car light comprises:

[0085] H1: using the vehicle driving environment perception method of embodiment one to perceive the environment in front of the vehicle to obtain a perception result, the perception result including a set target and its pixel coordinate position within the to-be-detected image, and the distance from the set target to the binocular camera;

[0086] H2: establishing a correspondence between the set target and the light area of the car light based on the optical characteristics of the DLP intelligent car light, and determining the area that needs to be adjusted in brightness; associating the identified set target with the illumination area of the car light to facilitate subsequent dynamic adjustment of the light beam distribution of the car light; specifically, using the target pixel coordinates and distance information obtained in step H1 to determine the position of the target object in the actual space, and combining the light area correspondence in step H2 to determine the light area range where the target is located, i.e. the area that needs to be shaded; in the implementation process, the shading area needs to be updated in real time according to the movement of the target object, such as the movement of oncoming vehicles and pedestrians, and the shading area needs to change with the vehicle speed and the relative speed of the target to ensure that it always covers the position that may cause glare. This updating process is realized based on real-time perception of the vehicle driving environment;

[0087] H3: independently controlling a plurality of micromirrors through a DMD control chip to adjust the brightness of each area.

[0088] The DMD control chip of the DLP intelligent car light divides the coverage area of the white laser light source into millions of independently controllable pixel points. The DMD chip is composed of millions of independently controllable micromirrors. Each micromirror can quickly switch between + / - 12 degrees. The brightness adjustment is realized by controlling the flipping timing of the micromirror.

[0089] Through the above control, the DLP intelligent car light can perform pixel-level accurate control on the illumination area, avoid the limitations of traditional mechanical light shielding, have fast response speed, and can respond to changes in road conditions in real time, while ensuring safe driving and providing the best lighting effect.

[0090] As a preferred embodiment of the above, the determination of the area requiring brightness adjustment is combined with the determination of the ambient light intensity.

[0091] As a further optimization of step H3, the DMD control chip is used to independently control a plurality of micromirrors to adjust the brightness of each area, as shown in Figure 4 , including:

[0092] H31: Calculate the brightness requirement of each light zone according to the perception result and the ambient light intensity;

[0093] H32: Convert the brightness requirement of each light zone into a corresponding PWM signal to generate a duty cycle;

[0094] H33: Drive the micromirror switch through the PWM signal to adjust the light zone brightness.

[0095] In the above process, the brightness requirement of each light zone is accurately calculated according to the perception result and the ambient light intensity, which can effectively avoid glare interference to oncoming vehicles, pedestrians or other road users. The accurate dimming range can avoid unnecessary over-illumination, while ensuring a clear view for the driver and improving the safety of night driving. By analyzing the ambient light intensity and the target detection result in real time, the system can dynamically adjust the light zone brightness to adapt to changes in road conditions and lighting conditions. In the implementation process, the timing control is coordinated by an accurate timing controller to coordinate the actions of all micromirrors. Usually, a dedicated ASIC or FPGA is used to achieve this, and the refresh rate needs to reach more than 60Hz to avoid flickering. The micromirrors in the same area work based on a unified duty cycle to ensure uniform light zone brightness. The micromirrors in different areas work independently with their own duty cycles while maintaining the overall synchronization of the PWM period.

[0096] As a preferred embodiment of the above, the ambient light intensity is detected by a light sensor and / or identified from the image to be detected collected by a binocular camera.

[0097] In the preferred embodiment, the light sensor is specifically used to measure the intensity of ambient light, usually the overall brightness value in lux, which can quickly and directly provide the numerical value of the current ambient light intensity, independent of other sensing systems, and can independently complete the ambient light measurement; the image obtained by the binocular camera can be indirectly deduced by image processing algorithm to obtain the ambient light intensity; in the preferred embodiment, the above two methods can be used simultaneously and mutually assisted, or can be used alternatively, both of which are within the protection scope of the present application.

[0098] As a specific way, as shown in Figure 5 When the ambient light intensity is identified by the detected image collected by the binocular camera, it includes:

[0099] B1: Collecting left and right images by the binocular camera and performing time and space alignment; thus generating a pair of views to ensure that the scenes captured by the two images are consistent; in the specific implementation process, polar correction can be performed to adjust the corresponding scene points in the left and right images to the same row, ensuring that the parallax only exists in the horizontal direction during stereo matching; in order to ensure the implementation effect, distortion correction can be performed on the left and right images to remove the deformation caused by the optical lens;

[0100] B2: Obtaining the set target and its pixel coordinate position detected in the detected image by using the target detection model, and excluding the set target from the left and right images;

[0101] Excluding the local brightness abnormal area can avoid its interference with the evaluation of the ambient light intensity; by means of the results of the foregoing steps, additional costs can be avoided, and misjudgment caused by dynamic targets such as moving vehicles or pedestrians and abnormal light sources can be prevented;

[0102] B3: Converting the remaining part of the left and right images after exclusion into a gray image; after excluding the local interference area, the brightness information is extracted from the remaining part of the image to ensure that the evaluation result can better reflect the true ambient light intensity; specifically, the left and right images after excluding the target are converted into a gray image, the brightness component is extracted, and the interference of color information is removed; as a preferred way, the gray image can be divided into a plurality of small blocks, such as 8x8 or 16x16 pixels; the block can localize the brightness analysis and reduce the influence of single-point brightness change on global judgment;

[0103] B4: Calculating the average gray value as the current ambient light intensity. Taking the above block implementation as an example, the average gray value of each block can be calculated to obtain the brightness information of each block, and then the average value is calculated; after excluding the set target, the global brightness statistical result can better reflect the true ambient light intensity, and the global brightness intensity can be used as the core parameter of the vehicle lamp adjustment to provide a reliable basis for the brightness control of the intelligent vehicle lamp.

[0104] In which, the set target is eliminated from the left and right images, specifically:

[0105] The pixel value of the set target corresponding area is replaced with a set value, which can be the average gray value of the whole image, or other specific values; or, the set target corresponding area is marked as invalid and does not participate in the brightness calculation.

[0106] As another embodiment, when the ambient light intensity is detected by both the light sensor and the recognition of the detected image collected by the binocular camera, it includes:

[0107] C1: based on the detection result of the light sensor, when the difference between the current and previous detection results is greater than a set value, start the recognition of the detected image;

[0108] C2: revise the basis according to the recognition result of the detected image to obtain the final ambient light intensity.

[0109] In the above process, image recognition is triggered as needed, which can reduce power consumption and computing resource occupation, flexibly cope with complex lighting conditions, and improve system reliability with dual data sources. As a specific way to revise the basis according to the recognition result of the detected image, the weight corresponding to the recognition result of the detected image can be obtained through the average gray value obtained in step B4 of the above embodiment, and the weight corresponding to the detection result of the light sensor can be calculated, and the detection result of the light sensor is revised, and the efficient fusion of the two detection methods can be realized through the way of assigning weights and summing, which can fully exert the global detection capability of the light sensor and the local analysis capability of the binocular camera, and provide reliable guarantee for accurate evaluation of ambient light intensity. In the implementation process, since the gray value of the image reflects the brightness of the light, but the unit is different from the lux value detected by the light sensor, the gray value can be converted into a weight factor comparable to the lux value through a mapping function.

[0110] The following is an example of revision by weight, and other ways to achieve the above revision purpose are also within the protection scope of the present application:

[0111] L final =W sensor ·L sensor +W camera ·L cauera (1);

[0112] W sensor =1-W camera (2);

[0113] Wherein, L final is the final revised ambient light intensity value;

[0114] L sensora light intensity value detected by the light sensor;

[0115] L camera a light intensity value obtained by mapping the image gray value;

[0116] W sensor and W camera are weight factors corresponding to the detection result of the light sensor and the recognition result of the to-be-detected image, respectively.

[0117] Alternatively, as another way of jointly participating in the confirmation of the ambient light intensity by the light sensor and the binocular camera, the proportion of the excluded area in the left and right images to the total area of the left and right images can be calculated, and the size of W sensor and W camera in the above formula (1) is adjusted, that is, the size of W camera is increased when the exclusion proportion is high, so as to more accurately fuse the two data sources and obtain a more reliable ambient light intensity evaluation result.

[0118] In the implementation process, a dynamic environment classification and light intensity change compensation strategy can also be performed. Specifically, the ambient light intensity scene is dynamically classified according to the exclusion area proportion:

[0119] Low exclusion proportion (<10%): The scene illumination is uniform, and the detection result of the light sensor can be directly used;

[0120] Medium exclusion proportion (10%50%): Local brightness is abnormal, but the proportion is not high. The light sensor result is revised by slight adjustment of the weight factor, for example, the decrease amplitude of W sensor is controlled to be 10%~20%;

[0121] High exclusion proportion (>50%): Local brightness occupies a large area. The light sensor result is revised by large adjustment of the weight factor, for example, the decrease amplitude of W sensor is controlled to be 20%~50%.

[0122] The basic principles, main features and advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. Anti-glare control method for DLP smart car lights, including: The vehicle driving environment perception method described below is used to perceive the environment in front of the vehicle to obtain a perception result: The vehicle driving environment perception method is implemented based on a binocular camera, which is installed at a set position on the vehicle body. The method includes: using the binocular camera to collect road condition information within the shooting range, and transmitting the collection results as the image to be detected to a target detection model deployed on the vehicle computer in real time; using the target detection model to predict the image to be detected to detect the set target and its pixel coordinate position in the image to be detected; using a ranging algorithm to calculate the distance from the set target in the image to be detected to the binocular camera; the perception result includes the set target and its pixel coordinate position in the image to be detected, as well as the distance from the set target to the binocular camera; Based on the optical characteristics of the DLP smart headlight, a correspondence between the set target and the headlight area is established to determine the area where brightness adjustment is required; The DMD control chip is used to independently control several micromirrors to adjust the brightness of each area; It is characterized by: In the process of determining the area that needs brightness adjustment, the ambient light intensity is combined with the ambient light intensity, which is identified by the image to be detected captured by the binocular camera, including: The left and right images are captured by a binocular camera and aligned in time and space. The set target and its pixel coordinate position detected in the image to be detected using the target detection model are obtained, and the set target is removed from the left and right images. The remaining parts of the left and right images after removal are converted into grayscale images. The average grayscale value is calculated as the current ambient light intensity.

2. The anti-glare control method for DLP intelligent vehicle lights according to claim 1, characterized in that: The steps of implementing the ranging algorithm include: For each pixel of the image to be detected, the cost value of pixel matching is obtained by fusing the BT cost and the cost processed by the SobelX operator; Locally smoothing the cost value of each pixel to obtain a cost matrix; performing multi-path constraint cost aggregation on the cost matrix to obtain an aggregated cost matrix; Determining an optimal disparity value for each pixel based on the aggregated cost matrix, and obtaining a disparity map through post-processing optimization; Distance information is calculated based on the disparity map.

3. The anti-glare control method for DLP intelligent vehicle lights according to claim 2, characterized in that: The steps for calculating the BT cost include: Calculate the pixel (x R -n,y) and (x R +n,y) between sub-pixel positions (x R +x,y) grayscale value x∈[-n,n]; Calculate the right center of the input right eye image (x R +dn,y) and (x R +d+n,y) between the sub-pixel positions (x R +d+x,y) grayscale value x∈[-n,n]; The BT price is calculated by the following formula: cos = min(cos1, cos2); Where cos is the BT cost, cos1 and cos2 are the bidirectional cost values, d is the disparity, 0.1≤n≤0.9, I R (x R ,y) is the sub-pixel position in the left image (x R ,y) at the gray value, I T (x R +d,y) is the sub-pixel position (x + d,y) in the right image that is offset by the parallax d. R +d,y) the grayscale value at that location.

4. The anti-glare control method for DLP intelligent vehicle lights according to claim 1, characterized in that: The DMD control chip independently controls several micromirrors to adjust the brightness of each area, including: Calculating the brightness requirement of each light zone based on the perception result and the ambient light intensity; Convert the brightness requirement of each light zone into a corresponding PWM signal to generate a duty cycle; The PWM signal drives the micromirror switch to adjust the brightness of the light zone.

5. The anti-glare control method for DLP intelligent vehicle lights according to claim 1, characterized in that: The set target is removed from the left and right images, specifically: Replacing the pixel value of the area corresponding to the set target with the set value; Alternatively, the area corresponding to the set target is marked as invalid and does not participate in the brightness calculation.

Images