Anti-glare dynamic dimming control system and method based on ambient light sensing

Abstract

The present invention discloses an anti-glare dynamic dimming control system and method based on ambient light sensing, specifically relating to the technical field of dimming control. The present invention captures the environmental picture in front of the vehicle lamp, obtains the video stream collected by the camera, processes the picture, and then judges the type of the target in the picture and its pixel position. According to the installation height of the camera, the vertical pixel ordinate of the target bottom and the vertical pixel offset of the reference line in the image relative to the horizontal axis, combined with the focal length of the camera, the threshold is calculated and adjusted. When the adjusted threshold exceeds the preset threshold, the ground hypothesis method constraint formula is used to calculate the depth information of the target. According to the type of the target and its position information, for the normal lighting picture, an anti-glare dark block is generated. The size and darkness of the generated dark block will be dynamically adjusted according to the difference between the illumination intensity of the current headlamp and the ambient illumination intensity and the target depth information.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic dimming, and more specifically, to an anti-glare dynamic dimming control system and method based on ambient light sensing. Background Art

[0002] With the acceleration of the urbanization process and the development of technology, artificial light sources have been widely used in various places. Due to inappropriate light intensity and direction, light pollution and glare phenomena are caused, affecting people's quality of life and work. In road lighting, strong light sources can cause discomfort to the eyes, even affect eyesight, cause visual fatigue, and lead to accidents. Therefore, solving these problems is of great significance for improving environmental comfort and visual health.

[0003] In the prior art, when the road surface is uneven, the calculated target depth information is inaccurate, which affects the generation of subsequent anti-glare dark blocks, resulting in poor anti-glare effects; and the anti-glare dark blocks with fixed size and darkness cannot adapt to different environments. To solve the above two defects, a technical solution is provided herein. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an anti-glare dynamic dimming control system and method based on ambient light sensing to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An anti-glare dynamic dimming control method based on ambient light sensing, comprising the following steps:

[0007] Capture the environmental picture in front of the vehicle headlight, obtain the video stream captured by the camera, process the picture and judge the target type and target position information of the picture;

[0008] Calculate the adjustment threshold according to the ratio of the camera installation height to the vertical pixel ordinate of the target bottom and the vertical pixel offset of the reference line in the image relative to the horizontal axis with respect to the camera focal length. When the adjustment threshold is greater than the preset threshold, introduce the ground hypothesis method constraint formula to calculate the target depth information;

[0009] According to the target type and target position information, for the normal lighting picture, generate anti-glare dark blocks, and the size and darkness of the generated dark blocks are dynamically controlled according to the difference between the current headlight illumination intensity and the ambient light illumination intensity and the target depth information.

[0010] In a preferred embodiment, YOLOv5+StrongSORT is used to obtain the video stream captured by the camera, process the picture, judge the type of the object in the picture and the pixel position of the object, and continuously track the object through the StrongSORT algorithm.

[0011] In a preferred embodiment, based on a known set of true heights and measured corresponding pixel data clues, monocular depth estimation is performed; the following formula is used to calculate the target depth information, and the formula is as follows: ; where f represents the camera focal length; represents the true physical height of the target object; represents the pixel height of the target in the image.

[0012] In a preferred embodiment, the ratio of the camera mounting height to the vertical pixel coordinate of the target bottom and the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis are determined; the distance between the vehicle and the target is deduced through the ratio of the camera mounting height to the vertical pixel coordinate of the target bottom; the vehicle tilt situation is calculated through the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis, and the specific formula is as follows: ; f represents the camera focal length; represents the vertical pixel offset of the reference line in the image relative to the horizontal axis.

[0013] In a preferred embodiment, the ratio of the camera mounting height to the vertical pixel coordinate of the target bottom and θ are determined, and the adjustment threshold is calculated by weighted summation. When the adjustment threshold is greater than the system preset threshold, the ground hypothesis method constraint formula is introduced to calculate the target depth information, and the formula is as follows: ; represents the camera mounting height; f represents the camera focal length; v represents the vertical pixel coordinate of the target bottom; represents the vertical coordinate of the image principal point.

[0014] In a preferred embodiment, according to the target type and target position information, for the normal lighting picture, an anti-glare dark block is generated, the difference between the illumination intensity of the current headlight and the ambient illumination intensity and the target depth information are determined, and the dark block adjustment coefficient is calculated by weighted summation.

[0015] In a preferred embodiment, the darkness of the dark block is adjusted according to the dark block adjustment coefficient; the size of the dark block is negatively correlated with the target depth information.

[0016] In a preferred embodiment, the anti-glare dynamic dimming control system based on ambient light sensing includes the following modules: video capture preprocessing module, depth information calculation module, projection module;

[0017] The video capture preprocessing module is used to capture the environmental picture in front of the vehicle lamp; then it uses YOLOv5+StrongSORT to obtain the video stream captured by the camera, processes the picture, judges the types of objects in the picture and the pixel positions of the objects, outputs the object bounding boxes to enclose the objects; and continuously tracks the objects through the StrongSORT algorithm;

[0018] The depth information calculation module is used to perform monocular depth estimation by based on a known set of true heights and measured corresponding pixel data clues; and calculate the target depth information;

[0019] The depth information calculation module internally includes a decision module, which is used to judge whether to introduce the ground hypothesis method constraint to calculate the target depth information by the ratio of the camera installation height to the vertical pixel ordinate of the target bottom and the vertical pixel offset of the reference line in the image relative to the horizontal axis and the camera focal length.

[0020] The projection module is used to generate anti-glare dark blocks for the normal lighting picture according to the object type and object position information, and the size and darkness of the generated dark blocks are dynamically controlled according to the difference between the current headlamp illumination intensity and the ambient illumination intensity and the target depth information.

[0021] In a preferred embodiment, the projection module internally includes an adjustment module, which is used to dynamically control the size and darkness of the generated dark blocks according to the difference between the current headlamp illumination intensity and the ambient illumination intensity and the target depth information.

[0022] The technical effects and advantages of the present invention:

[0023] The present invention first captures the environmental picture in front of the vehicle lamp through a roof camera; then uses YOLOv5+StrongSORT to obtain the video stream captured by the camera, processes the picture, judges the types of objects in the picture and the pixel positions of the objects, outputs the object bounding boxes to enclose the objects, and continuously tracks the objects through the StrongSORT algorithm; outputs specific numbers and the object bounding boxes and their trajectory information under the same number; performs monocular depth estimation based on a known set of true heights and measured corresponding pixel data clues; calculates the camera focal length through the known information, and combines the typical height information of vehicles, pedestrians, bicycles, and motorcycles to calculate the depth information of important objects in the current picture. When the ground is uneven, the measured height information is inaccurate, which affects the calculation of the depth information; the ground hypothesis method is introduced based on the similar triangle method for calculation, and it is comprehensively judged whether to introduce the ground hypothesis method to calculate the object depth information according to the ratio of the camera mounting height to the vertical pixel ordinate of the object bottom and the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis. The ground hypothesis method uses the camera mounting height and the vertical pixel ordinate of the object bottom to deduce the object depth information, making the calculation of the object depth information more accurate, applicable to complex terrains, and facilitating the subsequent generation of dark blocks.

[0024] Based on the three-dimensional information of the object, namely the two-dimensional pixel coordinates and the depth information, substitute them into the formula for converting the pixel coordinate system to the camera coordinate system to obtain the three-dimensional spatial position information of the object relative to the camera coordinate system; according to the object type and the object position information, determine the projection mode and the projection strategy. That is, for the normal lighting picture, generate anti-glare dark blocks to block the upper body of the pedestrian and avoid direct projection, achieving the anti-glare effect. The size and darkness of the generated dark blocks can be dynamically controlled according to the difference between the illumination intensity of the current headlight and the ambient illumination intensity and the object depth information. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings;

[0026] Figure 1 is a schematic flow chart of the anti-glare dynamic dimming control method based on ambient light sensing of the present invention;

[0027] Figure 2 is a schematic structural diagram of the anti-glare dynamic dimming control system based on ambient light sensing of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0028] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0029] Embodiment 1

[0030] The anti-glare dynamic dimming control method based on ambient light sensing of the present invention, as Figure 1 shown, includes the following steps:

[0031] First, capture the environmental picture in front of the vehicle headlight through the roof camera; determine that the picture captured by the camera contains the influence range of the vehicle headlight;

[0032] Then use YOLOv5+StrongSORT to obtain the video stream captured by the camera, process the picture and judge the type of the picture target and the pixel position where the target is located, output the target box to frame the target. And continuously track the target through the StrongSORT algorithm; output a specific number and the target box and its trajectory information under the same number;

[0033] Perform monocular depth estimation based on a known set of true heights and the measured corresponding pixel data clues; calculate the camera focal length through the known information, and combine the typical height information of vehicles, pedestrians, bicycles and motorcycles to calculate the target depth information.

[0034] When the ground is uneven, the measured height information is inaccurate, which affects the calculation of the depth information; the ground hypothesis method is introduced based on the similar triangle method for calculation, and it is comprehensively judged whether to introduce the ground hypothesis method to calculate the target depth information according to the ratio of the camera installation height and the vertical pixel coordinate of the target bottom and the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis. The ground hypothesis method uses the camera installation height and the vertical pixel coordinate of the target bottom to deduce the target depth information. This makes the calculation of the target depth information more accurate, applicable to complex terrains, and convenient for the subsequent generation of dark blocks.

[0035] Based on the three-dimensional information of the target, that is, the two-dimensional pixel coordinates and the depth information, substitute them into the formula for converting the pixel coordinate system into the camera coordinate system to obtain the three-dimensional spatial position information of the target relative to the camera coordinate system;

[0036] According to the target type and target position information, determine the projection mode and projection strategy. That is, for the normal lighting image, generate anti-glare dark blocks to block the upper body of pedestrians and avoid direct projection, so as to achieve the anti-glare effect. The size and darkness of the generated dark blocks can be dynamically controlled according to the difference between the illumination intensity of the current headlight and the ambient illumination intensity, and the target depth information. Among them, the greater the difference between the illumination intensity of the headlight and the ambient illumination intensity and the smaller the target depth information, the darker the darkness of the dark block, and the size of the dark block is negatively correlated with the target depth information.

[0037] Specifically,

[0038] First, capture the environmental image in front of the vehicle lights through the roof camera; determine that the captured image by the camera contains the influence range of the vehicle headlights; avoid dead corners that affect the acquisition of the video stream.

[0039] Then use YOLOv5+StrongSORT to obtain the video stream captured by the camera. YOLOv5 is a popular object detection algorithm, which is an iterative version in the YOLO series. The design concept of YOLOv5 is to provide high accuracy while maintaining a high detection speed. It is widely used in various computer vision tasks. The network architecture of YOLOv5 includes multiple versions, such as YOLOv5s (small), YOLOv5m (medium), YOLOv5l (large), and YOLOv5x (extra-large), to adapt to different performance requirements and resource limitations. These versions provide different trade-offs between model size, speed, and accuracy.

[0040] The network structure of YOLOv5 is the basis for its efficient object detection performance. This architecture inherits and develops the design concept of the YOLO series, aiming to quickly and accurately detect multiple objects in an image through a single forward propagation. The network structure of YOLOv5 contains four key parts: the input layer, the backbone network, the neck, and the detection head.

[0041] Input layer: The input layer is responsible for receiving the input image and performing preprocessing, including resizing the image to meet the size requirements of the network. YOLOv5 supports multi-scale input, which means it can process images of different resolutions, thereby improving the adaptability and detection efficiency of the model.

[0042] Backbone network: The backbone network is the core part of the model, responsible for extracting high-level features from the input image. YOLOv5 uses CSPDarknet53 as its backbone network, which is a deep convolutional neural network. By introducing the CrossStagePartialNetwork structure, it reduces the computational amount and improves the efficiency of the information flow. CSPDarknet53 contains multiple convolutional layers, residual connections, and downsampling operations, designed to capture rich features from shallow to deep layers.

[0043] Neck: The neck part located between the backbone network and the detection head is responsible for deeply integrating and processing feature information at different levels. To achieve this, YOLOv5 uses the structural design of the Feature Pyramid Network (FPN) and the Path Aggregation Network (PAN). FPN enhances the transmission of high-level semantic features at different scales through a top-down path, while PAN strengthens the utilization of low-level features through a bottom-up path. This two-way feature fusion strategy enables the model to effectively detect objects of different sizes.

[0044] Detection Head: The detection head is the last part of the YOLOv5 network structure and is responsible for generating the final detection results based on the fused feature maps. This includes the location (bounding box), class, and confidence of the object. The detection head of YOLOv5 uses multiple parallel convolutional layers to predict object information at different scales, and each scale of prediction is processed by a specific convolutional layer, thus achieving a wide range of detections from small objects to large objects.

[0045] Generally speaking, the network structure of YOLOv5 realizes the fast and accurate detection of objects through a carefully designed backbone network, an efficient feature fusion strategy, and multi-scale detection capabilities. The optimization of this structure aims to balance detection performance and computational efficiency, making YOLOv5 applicable not only to high-performance servers but also to effectively run on edge devices.

[0046] The loss function of YOLOv5 is the core of its learning process and is carefully designed to ensure high performance of the model in object detection tasks. This function combines three key parts: the localization loss of the bounding box, the classification loss of the object, and the confidence loss. These parts are elaborated in detail below:

[0047] Bounding Box Localization Loss: The bounding box localization loss uses the CIoU loss. This method not only focuses on the position and size accuracy of the bounding box but also comprehensively considers the consistency of the bounding box shape and the distance difference of the center points. The expression of the CIoU loss is as follows: ;

[0048] where IoU measures the overlap degree between the predicted box and the ground truth box, is the Euclidean distance between the center points of the predicted box and the ground truth box, c is the length of the diagonal of the smallest closed region covering the two boxes, v considers the consistency of the aspect ratio, and α is a tuning parameter used to balance the influence of the aspect ratio. This composite loss ensures that the model comprehensively considers the size, shape, and position of the bounding box during the learning process.

[0049] Object Classification Loss: The object classification loss uses the cross-entropy loss function, which is the standard method for evaluating the consistency between the predicted class of the model and the true class. Its purpose is to minimize the difference between the predicted class distribution and the true class label, and the formula is as follows: ;

[0050] Here, represents the one - hot encoded vector of the true class, while represents the class probabilities predicted by the model. By optimizing this loss, YOLOv5 effectively improves the accuracy of recognizing various targets.

[0051] Confidence loss: The confidence loss focuses on the prediction accuracy of whether the model detects the presence of an object and uses binary cross - entropy to quantify the difference between the predicted confidence and the actual situation. The specific expression is: ; Here, y is the true value indicating the presence or absence of the annotated object (1 for presence, otherwise 0), is the probability of presence predicted by the model. This part of the loss helps the model more accurately distinguish between the background and the foreground, reducing false detections and missed detections.

[0052] In summary, through a carefully designed combination, the loss function of YOLOv5 enables the model to achieve an excellent balance in object detection, classification, and confidence prediction, which is the key to its efficient and accurate object detection in diverse and complex environments.

[0053] The core technical architecture of the SORT method is based on the collaborative mechanism of motion state prediction and data association. Specifically, the Kalman filter dynamically predicts the motion trajectory of an object between consecutive video frames by establishing an object motion model; at the same time, the Hungarian algorithm, as an optimization and matching tool, is responsible for optimally associating the predicted trajectory with the detected object bounding boxes in the current frame. The limitations of this method are mainly reflected in the single - dimensional feature modeling. Since it does not integrate apparent feature information such as the texture and color of the object, when there are complex situations such as object occlusion and dense cross - motion in the scene, the tracking system is prone to problems such as trajectory breakage or identity label confusion.

[0054] DeepSORT is a deep - learning - based multi - object tracking algorithm widely used in fields such as video surveillance, autonomous driving, and behavior analysis. DeepSORT is an extension of the SORT algorithm; DeepSORT introduces appearance features on the basis of SORT and extracts the appearance information of the object through a deep - learning model, significantly reducing the ID switching problem caused by occlusion or temporary disappearance.

[0055] Predict the target position using the Kalman filter and measure the similarity between the prediction and the detection box using the Mahalanobis distance. Combine the cosine similarity of the deep feature vectors to ensure the consistency of the target identity. Improve the matching robustness by weighted fusion of motion and appearance information (such as 0.6 weight for appearance).

[0056] Prioritize matching the most recently appeared targets to avoid misassociating newly detected targets with long-lost tracks, thus solving the tracking problem in frequently occluded scenarios.

[0057] StrongSORT further improves the feature extraction and data association parts of DeepSORT. It adopts a more advanced feature extraction network and a more complex data association strategy to improve the accuracy and robustness of tracking. StrongSORT particularly enhances the ability to handle occlusions and interactions in complex scenarios.

[0058] Optimization algorithm for the appearance feature branch: In terms of the appearance feature branch of the StrongSORT algorithm, two key optimizations are mainly implemented: First, the performance of the feature extractor is enhanced; second, the exponential moving average (EMA) is adopted as the feature update mechanism to replace the feature library method used in the DeepSORT algorithm.

[0059] Optimization algorithm for the motion model branch: In the motion model branch of the StrongSORT algorithm, two key optimizations are mainly carried out: First, the compensation for camera motion is achieved by adopting the enhanced correlation coefficient (ECC) algorithm; second, the NSA (nonlinear robust adaptive) Kalman filter algorithm is introduced to replace the traditional Kalman filter algorithm, thus improving the accuracy and stability of tracking.

[0060] StrongSORT model construction: First, the detection results of the YOLOv5s algorithm provide the initial input for the entire tracking process. Based on the output of the current video frame, the StrongSORT algorithm generates a list containing all detected targets. Then, using the NSA Kalman filter algorithm, according to the tracking data of the previous frame, the possible positions and states of each target in the new frame are predicted. To match the tracking data with the new detection results, the algorithm combines the appearance features of the targets and adopts a global linear matching mechanism based on the Hungarian algorithm for association. When the initial association is not completely successful, the algorithm further adopts the IOU matching strategy to increase the matching success rate of the tracking targets. This series of operations ensures that the algorithm can identify which tracking targets match the detection results, which ones are still unmatched, and identify new unmatched detected objects. After completing the matching process, the StrongSORT algorithm uses the NSA Kalman filter to update the tracking data again. This step is to optimize the tracking results and improve their accuracy and robustness in the current frame. The whole process not only ensures the continuity and stability of tracking but also enhances the algorithm's adaptability to occlusions and dynamic changes, thus achieving efficient and accurate target tracking in real-time tracking tasks.

[0061] Perform monocular depth estimation by relying on a known set of true heights and measured corresponding pixel data clues. Determine the camera focal length, and combine the typical height information of vehicles, pedestrians, bicycles, and motorcycles to calculate the target depth information.

[0062] Furthermore, use the following formula to calculate the target depth information, and the formula is as follows: ; where f represents the camera focal length, which can be calculated offline by the checkerboard calibration method; represents the true physical height of the target object, which is dynamically updated through a preset typical target height library (such as through continuous tracking and optimization); represents the pixel height of the target in the image, which is obtained by calculating the bounding box output by detection models such as YOLO.

[0063] Furthermore, the target needs to be upright and have a stable height (such as a pedestrian standing, a vehicle not overturned). If the target bends down or has a tilted posture, it needs to be corrected according to the joint angle (such as when bending down drops to 1.3m).

[0064] Determine the ratio of the camera installation height to the vertical pixel ordinate of the target bottom and the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis; roughly estimate the distance between the vehicle and the target through the ratio of the camera installation height to the vertical pixel ordinate of the target bottom; calculate the vehicle tilt situation through the vertical pixel offset of the camera focal length and the reference line in the image relative to the horizontal axis. The specific formula is as follows: ; f represents the camera focal length; represents the vertical pixel offset of the reference line in the image relative to the horizontal axis.

[0065] Furthermore, calculate the adjustment threshold G through weighted summation. The formula is as follows: G = c * BZ + d * θ; where BZ represents the ratio of the camera installation height to the vertical pixel ordinate of the target bottom; c and d are the weight coefficients of the ratio of the camera installation height to the vertical pixel ordinate of the target bottom and θ respectively, and both are greater than zero.

[0066] When G is greater than the system preset threshold, it means the ground is uneven, and the simple similar triangle method cannot accurately calculate the target depth information. It is necessary to introduce the ground hypothesis method to constrain the formula, and then calculate the target depth information. The formula is as follows: ; represents the camera installation height (vertical distance from the ground), which needs to be measured after installation and calibrated regularly (such as a change of ±0.1m due to tire wear); f represents the camera focal length, and the acquisition method is the same as above; v represents the vertical pixel ordinate of the target bottom, that is, the center point coordinate at the bottom of the target detection box; represents the vertical ordinate of the image principal point (optical axis center).

[0067] Further, the camera pitch angle is monitored in real time by an angle sensor. When the deviation of the camera pitch angle is 1°, the depth error at 10 m is about 0.3 m, and dynamic compensation with the IMU is required. On a bumpy road, The instantaneous change needs to be corrected by the motion model of the suspension system.

[0068] Based on the three-dimensional information of the target, that is, the two-dimensional pixel coordinates and depth information, substituting into the formula for converting the pixel coordinate system to the camera coordinate system, the three-dimensional spatial position information of the target relative to the camera coordinate system is obtained;

[0069] According to the target type and target position information, the projection mode and projection strategy are determined. That is, for the normal lighting image, an anti-glare dark block is generated to block the upper body of the pedestrian and avoid direct projection, achieving the anti-glare effect.

[0070] Further, the size and darkness of the generated dark block can be dynamically controlled according to the difference between the current headlight illumination intensity and the ambient illumination intensity and the target depth information.

[0071] Specifically, the dark block adjustment coefficient is calculated by weighted summation, and the formula is as follows: T = a*cz + b*Z; where T represents the dark block adjustment coefficient; cz represents the difference between the current headlight illumination intensity and the ambient illumination intensity; Z represents the target depth information; a and b are the weight coefficients of the difference between the current headlight illumination intensity and the ambient illumination intensity and the target depth information respectively; the greater the difference between the headlight illumination intensity and the ambient illumination intensity and the smaller the target depth information, the darker the dark block. The darkness of the dark block is adjusted according to the dark block adjustment coefficient; the size of the dark block is negatively correlated with the target depth information.

[0072] The current headlight illumination intensity can be obtained through a photodiode array integrated behind the headlight, and the ambient illumination intensity can be obtained through a light intensity sensor installed on the roof.

[0073] Embodiment 2

[0074] The anti-glare dynamic dimming control system based on ambient light sensing of the present invention, as Figure 2 shown, includes the following modules: a video capture preprocessing module, a depth information calculation module, and a projection module;

[0075] The video capture preprocessing module is used to capture the environmental image in front of the vehicle lamp; then YOLOv5 + StrongSORT is used to obtain the video stream captured by the camera, process the image, judge the type of the target in the image and the pixel position where the target is located, output the target box to frame the target; and continuously track the target through the StrongSORT algorithm;

[0076] The depth information calculation module is used to perform monocular depth estimation by based on a known set of true heights and measured corresponding pixel data clues, and calculate the target depth information.

[0077] The depth information calculation module internally includes a decision module, which is used to judge whether to introduce the ground hypothesis method constraint to calculate the depth information by the ratio of the camera installation height to the vertical pixel ordinate of the target bottom and the vertical pixel offset of the reference line in the image relative to the horizontal axis and the camera focal length.

[0078] The projection module is used to determine the projection mode and projection strategy according to the target type and target position information.

[0079] The projection module internally contains an adjustment module, which is used to dynamically control the size and darkness of the generated dark block according to the difference between the current headlight illumination intensity and the ambient illumination intensity and the target depth information.

[0080] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but this implementation should not be considered to exceed the scope of this application.

[0081] In several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, and the indirect coupling or communication connection of the device or unit may be in an electrical, mechanical or other form.

[0082] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0083] In addition, the functional units in each embodiment of this application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0084] As described above, it is only the specific implementation manner of the present application. However, the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. Anti-glare dynamic dimming control method based on ambient light sensing, characterized in that: The following steps are involved: Capture the environment in front of the headlights, obtain the video stream captured by the camera, process the image and determine the type and location of the target in the image; The adjustment threshold is calculated according to the ratio of the camera installation height and the vertical coordinate of the pixel at the bottom of the target, as well as the camera focal length and the vertical pixel offset of the reference line in the image relative to the horizontal axis. When the adjustment threshold is greater than the preset threshold, the ground assumption method constraint formula is introduced to calculate the target depth information; According to the target type and target position information, anti-glare dark blocks are generated for the normally lit picture. The size and darkness of the generated dark blocks are dynamically controlled according to the difference between the current headlight light intensity and the ambient light intensity and the target depth information.

2. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 1, characterized in that: Use YOLOv5+StrongSORT to obtain the video stream captured by the camera, process the picture and determine the type of target in the picture and the pixel position of the target, and continue to track the target through the StrongSORT algorithm.

3. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 1, characterized in that: Perform monocular depth estimation based on a known set of true heights and corresponding pixel data clues; Use the following formula to calculate the target depth information: ; Where f represents the focal length of the camera; Represents the actual physical height of the target object; Represents the pixel height of the target in the image.

4. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 1, characterized in that: Determine the ratio of the camera installation height and the target bottom pixel ordinate, as well as the camera focal length and the vertical pixel offset of the reference line in the image relative to the horizontal axis; calculate the distance between the vehicle and the target by the ratio of the camera installation height and the target bottom pixel ordinate; calculate the vehicle inclination by the camera focal length and the vertical pixel offset of the reference line in the image relative to the horizontal axis. The specific formula is as follows: ; f represents the focal length of the camera; Represents the vertical pixel offset of the reference line in the image relative to the horizontal axis.

5. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 4, characterized in that: Determine the ratio of the camera installation height and the target bottom pixel ordinate as well as θ, calculate the adjustment threshold by weighted summation, and when the adjustment threshold is greater than the system preset threshold, introduce the ground assumption constraint formula to calculate the target depth information. The formula is as follows: ; represents the camera installation height; f represents the camera focal length; v represents the vertical coordinate of the bottom pixel of the target; Represents the ordinate of the principal point of the image.

6. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 1, characterized in that: According to the target type and target position information, anti-glare dark blocks are generated for the normal lighting picture, the difference between the current headlight light intensity and the ambient light intensity and the target depth information are determined, and the dark block adjustment coefficient is calculated through weighted summation.

7. The anti-glare dynamic dimming control method based on ambient light sensing according to claim 6, characterized in that: The darkness of the dark block is adjusted according to the dark block adjustment coefficient; the size of the dark block is negatively correlated with the target depth information. 8.Anti-glare dynamic dimming control system based on ambient light sensing, characterized in that: It includes the following modules: video capture preprocessing module, depth information calculation module, and projection module; The video capture preprocessing module is used to capture the environment in front of the headlights; Then use YOLOv5+StrongSORT to obtain the video stream captured by the camera, process the image and determine the type and location of the target in the image, output the target frame, frame the target; and continue to track the target through the StrongSORT algorithm; The depth information calculation module is used to perform monocular depth estimation based on a known set of real heights and measured corresponding pixel data clues; calculate the target depth information; The depth information calculation module includes a decision module for determining whether to introduce a ground hypothesis constraint to calculate the target depth information based on the ratio of the camera installation height and the vertical coordinate of the pixel at the bottom of the target, the camera focal length, and the vertical pixel offset of the reference line in the image relative to the horizontal axis; The projection module is used to generate anti-glare dark blocks for the normally lit picture according to the target type and target position information. The size and darkness of the generated dark blocks are dynamically controlled according to the difference between the current headlight light intensity and the ambient light intensity and the target depth information.

9. The anti-glare dynamic dimming control system based on ambient light sensing according to claim 8, characterized in that: The projection module includes an adjustment module for dynamically controlling the size and darkness of the generated dark block according to the difference between the current headlight illumination intensity and the ambient illumination intensity and the target depth information.

Images