Floodlight lens illumination range optimization method and lens

Abstract

The present invention belongs to the field of artificial intelligence technology and discloses a method for optimizing the lighting range of a spotlight lens and a lens. The method comprises: constructing an environment in which a lamp-borne device is located into a three-dimensional environmental space; collecting detection target parameter information in real time; obtaining a detection target category based on the detection target parameter information; performing feature extraction on the detection target parameter information based on the detection target category to obtain detection target feature data; collecting lamp-borne device parameter information in real time; processing the detection target parameter information and lamp-borne device parameter information based on the detection target feature data to obtain a detection target moving speed and a detection target moving direction; adaptively adjusting the spotlight beam based on the lamp-borne device parameter information, the detection target category, the detection target feature data, the detection target moving speed and the detection target moving direction; the solution dynamically adjusts the lighting range according to changes in the scene and the detection target to meet the requirements of dynamic lighting.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and more particularly to a method for optimizing the illumination range of a floodlight lens and the lens. Background Art

[0002] Floodlights are essential equipment used in a wide range of applications, including road lighting, architectural projection, and scene rendering. Traditional floodlights typically utilize fixed beam angles and illumination ranges, meeting basic lighting requirements in some scenarios. However, with the increasing demand for energy conservation, environmental protection, precise lighting, and versatility in modern society, the technical limitations of floodlights are becoming increasingly apparent. Fixed illumination ranges are difficult to adapt to diverse scenarios, limiting their application scenarios and efficiency.

[0003] Traditional floodlights typically have a fixed illumination range and beam angle, making them difficult to adapt to different scene requirements. They also lack intelligent, real-time feedback mechanisms, making them incapable of meeting dynamic lighting requirements. Therefore, there is an urgent need for a floodlight lens illumination range optimization technology that can dynamically adjust according to real-time scene conditions to meet diverse lighting needs. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a method for optimizing the illumination range of a floodlight lens, comprising:

[0005] Constructing the environment where the light-carrying device is located into a three-dimensional environment space, wherein the light-carrying device is a device having a floodlight lens;

[0006] Real-time collection of detection target parameter information;

[0007] According to the detection target parameter information, the detection target category is obtained;

[0008] Extracting features of the detection target parameter information according to the detection target category to obtain detection target feature data;

[0009] Real-time collection of lamp-borne equipment parameter information;

[0010] Processing the detection target parameter information and the light-borne equipment parameter information according to the detection target feature data to obtain the detection target moving speed and detection target moving direction;

[0011] The floodlight beam is adaptively adjusted according to the parameter information of the lamp-mounted equipment, the detection target category, the detection target feature data, the detection target movement speed and the detection target movement direction.

[0012] Furthermore, the method for adaptively adjusting the light beam of a floodlight according to the parameter information of the onboard device, the detection target category, the detection target characteristic data, the detection target moving speed and the detection target moving direction includes:

[0013] If the detection target category is a pedestrian, the spotlight beam is adaptively adjusted according to the detection target feature data, detection target movement speed and detection target movement direction;

[0014] If the detection target category is a lamp-mounted device, the spotlight beam is adaptively adjusted according to the detection target lamp-mounted device based on the detection target feature data, the detection target movement speed and the detection target movement direction.

[0015] Furthermore, the method for adaptively adjusting the light beam of a floodlight for pedestrians includes:

[0016] Dynamically adjust the horizontal and vertical angle ranges of the light-mounted device based on the horizontal and vertical angle ranges of the detection target, so that the horizontal and vertical angle ranges of the light-mounted device cover the horizontal and vertical angle ranges of the detection target. If the original angle range already covers the detection target angle range, maintain the original setting without adjustment. The details are as follows:

[0017] If the lower limit of the horizontal angle range of the lamp-mounted device is less than or equal to the lower limit of the horizontal angle range of the detection target, and the upper limit of the horizontal angle range of the lamp-mounted device is greater than or equal to the upper limit of the horizontal angle range of the detection target, then the horizontal angle range of the lamp-mounted device does not need to be adjusted;

[0018] If the lower limit of the vertical angle range of the lamp-mounted device is less than or equal to the lower limit of the vertical angle range of the detection target, and the upper limit of the vertical angle range of the lamp-mounted device is greater than or equal to the upper limit of the vertical angle range of the detection target, then the vertical angle range of the lamp-mounted device does not need to be adjusted;

[0019] If neither the horizontal angle range nor the vertical angle range of the lamp-mounted device needs to be adjusted, the current process ends;

[0020] If the lower limit of the horizontal angle range of the lamp-mounted device is greater than the lower limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the lower limit of the horizontal angle range of the lamp-mounted device is the lower limit of the horizontal angle range of the detection target; if the upper limit of the horizontal angle range of the lamp-mounted device is less than the upper limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the upper limit of the horizontal angle range of the lamp-mounted device is the upper limit of the horizontal angle range of the detection target;

[0021] If the lower limit of the vertical angle range of the lamp-mounted equipment is greater than the lower limit of the vertical angle range of the detection target, the spotlight lens is adjusted so that the lower limit of the vertical angle range of the lamp-mounted equipment is the lower limit of the vertical angle range of the detection target; if the upper limit of the vertical angle range of the lamp-mounted equipment is less than the upper limit of the vertical angle range of the detection target, the spotlight lens is adjusted so that the upper limit of the vertical angle range of the lamp-mounted equipment is the upper limit of the vertical angle range of the detection target.

[0022] Furthermore, the method for adaptively adjusting the light beam of a floodlight according to the detection target lamp-mounted device includes:

[0023] Dynamically adjust the horizontal and vertical angle ranges of the light-borne equipment based on the horizontal and vertical angle ranges of the detection target, so that the horizontal angle range of the light-borne equipment covers the horizontal angle range of the detection target, and the lower limit of the vertical angle range of the light-borne equipment is not higher than the lower limit of the vertical angle range of the detection target;

[0024] Input the detection target feature data, detection target movement speed, and detection target movement direction into a pre-built angle setting model to obtain the upper limit of the vertical angle prediction of the lamp-mounted device, and dynamically adjust the upper limit of the vertical angle range of the lamp-mounted device based on the upper limit of the vertical angle prediction; the details are as follows:

[0025] If the lower limit of the horizontal angle range of the lamp-mounted device is less than or equal to the lower limit of the horizontal angle range of the detection target, and the upper limit of the horizontal angle range of the lamp-mounted device is greater than or equal to the upper limit of the horizontal angle range of the detection target, then the horizontal angle range of the lamp-mounted device does not need to be adjusted;

[0026] If the lower limit of the horizontal angle range of the lamp-mounted device is greater than the lower limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the lower limit of the horizontal angle range of the lamp-mounted device is the lower limit of the horizontal angle range of the detection target; if the upper limit of the horizontal angle range of the lamp-mounted device is less than the upper limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the upper limit of the horizontal angle range of the lamp-mounted device is the upper limit of the horizontal angle range of the detection target;

[0027] If the lower limit of the vertical angle range of the lamp-mounted device is less than or equal to the lower limit of the vertical angle range of the detection target, the lower limit of the vertical angle range of the lamp-mounted device does not need to be adjusted; if the lower limit of the vertical angle range of the lamp-mounted device is greater than the lower limit of the vertical angle range of the detection target, the lens of the floodlight is adjusted so that the lower limit of the vertical angle range of the lamp-mounted device is the lower limit of the vertical angle range of the detection target;

[0028] The detection target feature data, detection target movement speed and detection target movement direction are input into the pre-built angle setting model to obtain the predicted upper limit of the vertical angle of the lamp-mounted device; if the upper limit of the vertical angle range of the lamp-mounted device is equal to the predicted upper limit of the vertical angle of the lamp-mounted device, there is no need to adjust the upper limit of the vertical angle range of the lamp-mounted device; if the upper limit of the vertical angle range of the lamp-mounted device is not equal to the predicted upper limit of the vertical angle of the lamp-mounted device, the spotlight lens is adjusted so that the upper limit of the vertical angle range of the lamp-mounted device is the predicted upper limit of the vertical angle of the lamp-mounted device.

[0029] Furthermore, the parameter information of the lamp-mounted device includes a moving speed of the lamp-mounted device, a moving direction of the lamp-mounted device, a horizontal angle range of the lamp-mounted device, and a vertical angle range of the lamp-mounted device;

[0030] The method for acquiring the detection target moving speed and the detection target moving direction includes:

[0031] Obtain the center position of the detection target from the detection target feature data, obtain the three-dimensional coordinates of the light-carrying device from the three-dimensional environment space; obtain the distance between the detection target and the light-carrying device from the detection target parameter information;

[0032] The horizontal speed direction and the vertical speed direction of the detection target are calculated based on the center position of the detection target and the three-dimensional coordinates of the light-carrying device; the horizontal speed direction and the vertical speed direction of the detection target are constructed into the movement direction of the detection target;

[0033] Obtain the moving speed of the lamp-borne equipment from the lamp-borne equipment parameter information;

[0034] The moving speed of the detection target is calculated based on the moving direction of the detection target, the moving speed of the lamp-mounted device, the center position of the detection target, and the distance between the detection target and the lamp-mounted device.

[0035] Furthermore, the method for acquiring the detection target feature data includes:

[0036] Analyze and process the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height;

[0037] The center position of the detection target is calculated according to the target refinement starting position, the detection target refinement length, the detection target refinement width and the detection target refinement height;

[0038] The horizontal angle range and vertical angle range of the detection target are calculated according to the center position of the detection target, the refined width of the detection target and the refined height of the detection target;

[0039] The center position of the detection target, the horizontal angle range of the detection target and the vertical angle range of the detection target are constructed into the detection target feature data.

[0040] Furthermore, the method of analyzing and processing the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height includes:

[0041] Extract the starting position and spatial size based on the three-dimensional coordinate data of the detection target;

[0042] If the detection target is a detection target lamp-mounted device, based on the reflection intensity of the detection target, data points with reflection intensities within a preset glass reflection intensity range are extracted to form a glass area. Based on the glass area, the starting position and spatial dimensions are extracted as the refined features of the detection target, and the target refined starting position, detection target refined length, detection target refined width, and detection target refined height are obtained; specifically, as follows:

[0043] Obtain all three-dimensional coordinates of the detection target from the three-dimensional environment space, and record the smallest three-dimensional coordinate among all the three-dimensional coordinates of the detection target as the detection target starting position; obtain the detection target length, detection target width and detection target height from the detection target parameter information;

[0044] If the detection target category is a pedestrian, the detection target starting position is the target refinement starting position, the detection target length is the detection target refinement length, the detection target width is the detection target refinement width, and the detection target height is the detection target refinement height, and the current process ends;

[0045] If the detection target category is a lamp-mounted device, the detection target reflection intensity is obtained from the detection target parameter information; the area where the detection target reflection intensity is within the preset glass reflection intensity range is constructed as the lamp-mounted device glass area; the glass reflection intensity range is , is the lower limit of glass reflection intensity, is the upper limit of the glass reflection intensity, and are all constants;

[0046] The minimum three-dimensional coordinate within the glass area of the lamp-mounted equipment is used as the starting position for target refinement, the length of the glass area of the lamp-mounted equipment is used as the detection target refinement length, the width of the glass area of the lamp-mounted equipment is used as the detection target refinement width, and the height of the glass area of the lamp-mounted equipment is used as the detection target refinement height, and the current process ends.

[0047] Furthermore, the method for acquiring the detection target category includes:

[0048] Inputting the detection target parameter information into a pre-built category diagnosis model to obtain the detection target category, wherein the detection target category includes pedestrians and detection target light-mounted equipment;

[0049] The training method of the category diagnosis model includes:

[0050] Pre-collecting a category diagnosis dataset, the category diagnosis dataset including Q groups of category diagnosis data and detection target categories corresponding to the Q groups of category diagnosis data, where Q is a positive integer greater than 0, and the category diagnosis data including detection target parameter information; dividing the category diagnosis dataset into a training set and a validation set, wherein the training set is used to train a category diagnosis model, and the validation set is used to evaluate the generalization performance of the category diagnosis model;

[0051] During the training process of the category diagnosis model, minimizing the cross-entropy loss function is used as the optimization goal. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, training is stopped. The category diagnosis model is trained based on the logistic regression model or the support vector machine model.

[0052] The category diagnosis data is converted into a high-dimensional feature vector; the input layer of the category diagnosis model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through the hidden layer. Finally, the output layer of the category diagnosis model calculates the probability distribution of the detection target category through the softmax activation function, and outputs the detection target category corresponding to the maximum probability as the final prediction result.

[0053] Furthermore, the method for constructing the three-dimensional environment space includes:

[0054] The radar sensor on the light-mounted device scans the environment, collects the position and shape characteristics of objects in the environment, and generates raw point cloud data. The raw point cloud data consists of discrete data points, each of which corresponds to a location in the environment and records the three-dimensional coordinate information of the data point.

[0055] The collected original point cloud data is cleaned of noise, invalid data in the original point cloud data is removed, and valid data point information in the original point cloud data is retained to obtain the preprocessed environmental point cloud data;

[0056] Convert the pre-processed environmental point cloud data into an occupancy grid model, set the grid unit size for the occupancy grid model, divide the occupancy grid model into regular grid units with equal spacing according to the grid unit size; assign data points to corresponding regular grid units to complete the conversion from point data to unit data;

[0057] The regular grid cells occupied by obstacles are marked, and the marked regular grid cells are integrated into a three-dimensional map structure processed by a computer to obtain a three-dimensional environment space.

[0058] Furthermore, the training method of the angle setting model includes:

[0059] Preliminarily collecting an angle setting data set, the angle setting data set comprising R groups of angle setting data and upper limits of vertical angle prediction corresponding to the R groups of angle setting data, where R is a positive integer greater than 0, the angle setting data comprising detection target feature data, detection target movement speed, and detection target movement direction; dividing the angle setting data set into a training set and a validation set, wherein the training set is used to train an angle setting model, and the validation set is used to evaluate the generalization performance of the angle setting model;

[0060] During the training process of the angle setting model, the optimization objective is to minimize the cross entropy loss function. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, training is stopped. The angle setting model is trained based on the logistic regression model or the support vector machine model.

[0061] The angle setting data is converted into a high-dimensional feature vector; the input layer of the angle setting model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through the hidden layer. Finally, the output layer of the angle setting model calculates the probability distribution of the upper limit of the vertical angle prediction through the softmax activation function, and outputs the upper limit of the vertical angle prediction corresponding to the maximum probability as the final prediction result.

[0062] The projector lens includes: based on the projector lens illumination range optimization method, the projector beam is adaptively adjusted according to the lamp-borne equipment parameter information, the detection target category, the detection target characteristic data, the detection target movement speed and the detection target movement direction.

[0063] Compared with the prior art, the technical effects and advantages of the method for optimizing the illumination range of the floodlight lens and the lens of the present invention are as follows:

[0064] This solution acquires the target category by collecting target parameter information in real time. It adjusts the illumination range of the light beam according to the target category. It accurately calculates the horizontal angle range of the lighting equipment and the vertical angle range of the lighting equipment by detecting the target category, target feature data, target movement speed, and target movement direction. It also adaptively adjusts the floodlight beam. When the target moves or the light-mounted equipment moves, it adaptively adjusts according to different scene requirements to ensure that the illumination range matches the detection target requirements. This solution introduces dynamic lens adjustment technology to automatically adjust the illumination range of the floodlight lens. It can dynamically adjust the illumination range in real time according to changes in the scene and detection target, thereby achieving efficient, precise, and personalized lighting to meet the requirements of dynamic lighting. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 Schematic diagram of a system for optimizing the illumination range of a floodlight lens according to Example 1 of the present invention;

[0066] Figure 2 This is a flow chart of a method for optimizing the illumination range of a floodlight lens according to embodiment 2 of the present invention;

[0067] Figure 3 A flow chart of a method for adaptively adjusting a floodlight beam for pedestrians;

[0068] Figure 4 A flow chart of a method for adaptively adjusting a floodlight beam in response to a detection target lamp-mounted device. DETAILED DESCRIPTION

[0069] The technical solutions in the embodiments of the present invention will be described in detail, clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. It should be noted that the specific embodiments described below are only used to better illustrate and describe the technical solutions of the present invention, and are intended to enable those skilled in the art to better understand and implement the present invention, and should not be construed as limiting the scope of protection of the present invention. Without departing from the spirit and essence of the present invention, those skilled in the art may modify, adjust or make equivalent replacements based on the contents disclosed in the present invention, and these should all be regarded as the scope of protection of the present invention.

[0070] Example 1

[0071] See also Figure 1 As shown, the floodlight lens lighting range optimization system described in this embodiment includes a space construction module, a first acquisition module, a category diagnosis module, a first processing module, a second acquisition module, a second processing module and a beam adjustment module. Each module realizes data transmission through wired and / or wireless connections.

[0072] The space construction module is used to construct the environment where the lamp-carrying device is located into a three-dimensional environment space, and the lamp-carrying device is a device with a floodlight lens.

[0073] It should be noted that the light-carrying device is, for example, a vehicle.

[0074] The method for constructing the three-dimensional environment space includes:

[0075] The radar sensor on the light-mounted device scans the environment, collects the position and shape characteristics of objects in the environment, and generates raw point cloud data. The raw point cloud data consists of a large number of discrete data points, each of which corresponds to a location in the environment and records the three-dimensional coordinate information of the data point.

[0076] Perform noise cleaning on the collected raw point cloud data (such as outlier removal, ground point filtering, etc.), remove invalid data in the raw point cloud data, retain valid data point information in the raw point cloud data, and obtain pre-processed environmental point cloud data;

[0077] Convert the pre-processed environmental point cloud data into an occupancy grid model, set the grid unit size for the occupancy grid model, divide the occupancy grid model into regular grid units with equal spacing according to the grid unit size; assign data points to corresponding regular grid units to complete the conversion from point data to unit data;

[0078] The regular grid cells occupied by obstacles are marked, and the marked regular grid cells are integrated into a three-dimensional map structure that can be processed by a computer to obtain a three-dimensional environment space.

[0079] The first acquisition module is used to collect detection target parameter information in real time; the detection target parameter information includes the distance between the detection target and the lamp-mounted equipment, the detection target length, the detection target width, the detection target height and the detection target reflection intensity; the detection target parameter information is obtained through the radar sensor.

[0080] The category diagnosis module is used to obtain the category of the detection target based on the detection target parameter information; the detection target category includes pedestrians and detection target light-borne equipment.

[0081] The method for acquiring the detection target category includes:

[0082] Input the detection target parameter information into the pre-built category diagnosis model to obtain the detection target category;

[0083] The training method of the category diagnosis model includes:

[0084] Pre-collecting a category diagnosis dataset, the category diagnosis dataset including Q groups of category diagnosis data and detection target categories corresponding to the Q groups of category diagnosis data, where Q is a positive integer greater than 0, and the category diagnosis data including detection target parameter information; dividing the category diagnosis dataset into a training set and a validation set, wherein the training set is used to train a category diagnosis model, and the validation set is used to evaluate the generalization performance of the category diagnosis model;

[0085] During the training process of the category diagnosis model, minimizing the cross-entropy loss function is used as the optimization goal. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, training is stopped. The category diagnosis model is trained based on the logistic regression model or the support vector machine model.

[0086] The category diagnosis data is converted into a high-dimensional feature vector; the input layer of the category diagnosis model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through several hidden layers. Finally, the output layer of the category diagnosis model calculates the probability distribution of the detection target category through the softmax activation function, and outputs the detection target category corresponding to the maximum probability as the final prediction result.

[0087] It should be noted that different target categories require different lighting requirements. Classifying the detection targets helps to adaptively adjust the floodlight beam according to the actual detection target category, thereby achieving precise lighting, improving both safety and energy efficiency, ensuring the safety of drivers and pedestrians, and at the same time improving the intelligence level and energy-saving effect of the system.

[0088] For example, when the detection target category is a pedestrian, the light beam of the floodlight needs to be able to illuminate the entire height range of the pedestrian from feet to head. If the light beam does not cover the entire body, part of the pedestrian's body will be blocked or not illuminated, making it impossible for the driver to quickly judge the pedestrian's position, behavior and direction of travel, thereby increasing traffic risks.

[0089] For example, when the detection target category is a car, the car cab is located low, and the car driver's line of sight is about 1 to 1.2 meters from the ground. The beam of the floodlight should be slightly lower than the car driver's line of sight, and the beam angle should be relatively flat, mainly illuminating the front of the vehicle and the ground to prevent light from entering the cab.

[0090] For example, when the detection target category is a truck, the truck cab is located at a higher position, and the truck driver's line of sight is about 1.8 to 2.5 meters from the ground. The beam of the floodlight should be slightly lower than the truck driver's line of sight, and the beam angle should be larger, extending to a longer distance, taking into account the lighting needs of the truck's height and long body.

[0091] The first processing module is used to extract features of detection target parameter information according to the detection target category to obtain detection target feature data.

[0092] The method for acquiring the detection target feature data includes:

[0093] Analyze and process the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height;

[0094] The center position of the detection target is calculated according to the target refinement starting position, the detection target refinement length, the detection target refinement width and the detection target refinement height;

[0095] The horizontal angle range and vertical angle range of the detection target are calculated according to the center position of the detection target, the refined width of the detection target and the refined height of the detection target;

[0096] The center position of the detection target, the horizontal angle range of the detection target and the vertical angle range of the detection target are constructed into the detection target feature data.

[0097] The method of analyzing and processing the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height includes:

[0098] Extract the starting position and spatial size based on the three-dimensional coordinate data of the detection target;

[0099] If the detection target is a detection target lamp-mounted device, based on the reflection intensity of the detection target, data points with reflection intensities within a preset glass reflection intensity range are extracted to form a glass area. Based on the glass area, the starting position and spatial dimensions are extracted as the refined features of the detection target, and the target refined starting position, detection target refined length, detection target refined width, and detection target refined height are obtained; specifically, as follows:

[0100] Obtain all three-dimensional coordinates of the detection target from the three-dimensional environment space, and record the smallest three-dimensional coordinate among all the three-dimensional coordinates of the detection target as the detection target starting position; obtain the detection target length, detection target width and detection target height from the detection target parameter information;

[0101] If the detection target category is a pedestrian, the detection target starting position is the target refinement starting position, the detection target length is the detection target refinement length, the detection target width is the detection target refinement width, and the detection target height is the detection target refinement height, and the current process ends;

[0102] If the detection target category is a lamp-mounted device, the detection target reflection intensity is obtained from the detection target parameter information; the area where the detection target reflection intensity is within the preset glass reflection intensity range is constructed as the lamp-mounted device glass area; the glass reflection intensity range is , is the lower limit of glass reflection intensity, is the upper limit of the glass reflection intensity, and All are constants; the minimum three-dimensional coordinate within the lamp-mounted equipment glass area is used as the target refinement starting position, the length of the lamp-mounted equipment glass area is used as the detection target refinement length, the width of the lamp-mounted equipment glass area is used as the detection target refinement width, and the height of the lamp-mounted equipment glass area is used as the detection target refinement height, and the current process ends.

[0103] It should be noted that the target refinement starting position plays a decisive role in the subsequent adjustment of the floodlight beam. For pedestrians, directly using the starting position of the detection target as the target refinement starting position can take into account the actual position of the pedestrian's body in the calculation of the beam illumination range, ensuring that the pedestrian is illuminated while avoiding irradiation of other unrelated areas and resulting in waste of beam. For lamp-mounted equipment, considering the complexity of the lamp-mounted equipment, especially the reflective characteristics of the glass area, the glass area of the lamp-mounted equipment is screened out by detecting the target reflection intensity. The precise positioning of the glass area helps to adjust the illumination angle and range of the beam, ensuring that the light does not directly shine into the eyes of the lamp-mounted equipment driver, which not only improves the driving experience of the lamp-mounted equipment driver, but also reduces the risk of accidents caused by strong light stimulation.

[0104] The method for calculating the center position of the detection target according to the target refinement starting position, the detection target refinement length, the detection target refinement width, and the detection target refinement height includes:

[0105] ;

[0106] Will 、 and Constructed to detect the center position of the target ;

[0107] in, To detect the coordinate of the target center position on the X-axis, To detect the coordinate of the target center position on the Y axis, To detect the coordinate of the target center position on the Z axis, 、 and To detect the adjustment parameters of the target center position, Refine the coordinate of the starting position on the X axis for the target, Refine the coordinate of the starting position on the Y axis for the target, Refine the coordinates of the starting position on the Z axis for the target, To detect the target length, To refine the width of the detected target, Refine the height for detecting objects.

[0108] Methods for calculating the horizontal angle range of a detection target based on the detection target center position, the detection target refinement width, and the detection target refinement height include:

[0109] ;

[0110] ;

[0111] in, is the lower limit of the horizontal angle range of the detection target, is the upper limit of the horizontal angle range of the detection target, and is the adjustment parameter for the horizontal angle range of the detection target, is the inverse tangent function.

[0112] Methods for calculating the vertical angle range of a detection target based on the detection target center position, the detection target refinement width, and the detection target refinement height include:

[0113] ;

[0114] ;

[0115] in, is the lower limit of the vertical angle range of the detection target, is the upper limit of the vertical angle range of the detection target, and It is the adjustment parameter of the vertical angle range of the detection target.

[0116] It's important to note that the horizontal and vertical angle ranges of the detection target essentially describe the spotlight's field of view in the horizontal and vertical directions of the detection target, that is, the angle of the target's projection onto a plane. The target refinement width and height determine the coverage area of the spotlight's beam, while the target refinement length affects the distance of the spotlight's beam and does not directly affect the calculation of the horizontal and vertical angle ranges of the detection target.

[0117] The second acquisition module is used to collect parameter information of the lamp-borne equipment in real time; the parameter information of the lamp-borne equipment includes the moving speed of the lamp-borne equipment, the moving direction of the lamp-borne equipment, the horizontal angle range of the lamp-borne equipment and the vertical angle range of the lamp-borne equipment. The moving speed and moving direction of the lamp-borne equipment can be obtained through the global positioning system or the Beidou satellite navigation system. The floodlight is equipped with an automatic control system, and the horizontal angle range and the vertical angle range of the lamp-borne equipment are obtained through the automatic control system.

[0118] It should be noted that the movement direction of the light-borne equipment can be expressed in degrees, which are defined in a three-dimensional environment space as angles calculated clockwise from due north (for example, 0° represents north, 90° represents east, 180° represents south, and 270° represents west).

[0119] The second processing module is used to process the detection target parameter information and the lamp-borne device parameter information according to the detection target feature data to obtain the detection target movement speed and the detection target movement direction.

[0120] The method for acquiring the detection target moving speed and the detection target moving direction includes:

[0121] Obtain the center position of the detection target from the detection target feature data, obtain the three-dimensional coordinates of the light-carrying device from the three-dimensional environment space; obtain the distance between the detection target and the light-carrying device from the detection target parameter information;

[0122] The horizontal speed direction and the vertical speed direction of the detection target are calculated based on the center position of the detection target and the three-dimensional coordinates of the light-carrying device; the horizontal speed direction and the vertical speed direction of the detection target are constructed into the movement direction of the detection target;

[0123] Obtain the moving speed of the lamp-borne equipment from the lamp-borne equipment parameter information;

[0124] The moving speed of the detection target is calculated based on the moving direction of the detection target, the moving speed of the lamp-mounted device, the center position of the detection target, and the distance between the detection target and the lamp-mounted device.

[0125] The calculation method for detecting the horizontal speed direction of the target includes:

[0126] ;

[0127] in, To detect the horizontal speed direction of the target, is the angle calculation function, To detect the coordinate of the target center position on the X-axis, To detect the coordinate of the target center position on the Y axis, is the three-dimensional coordinate of the lamp-carrying equipment on the X-axis, is the three-dimensional coordinate of the lamp-carrying equipment on the Y axis, To detect the distance between the target and the light-borne equipment.

[0128] The calculation method for detecting the vertical velocity direction of the target includes:

[0129] ;

[0130] in, To detect the vertical velocity direction of the target, To detect the coordinate of the target center position on the Z axis, The three-dimensional coordinates of the lamp-mounted equipment on the Z axis.

[0131] It should be noted that calculating the horizontal and vertical velocity directions of the detection target can describe the target's relative position in the three-dimensional environment, ensuring that the lighting device's illumination range accurately covers the target. The horizontal velocity direction of the detection target represents the horizontal direction of the detection target relative to the lighting device (i.e., the angle between the horizontal velocity directions of the detection target and the lighting device). It is used to determine the left and right position of the detection target in front of the lighting device. The vertical velocity direction of the detection target represents the vertical direction of the detection target relative to the lighting device (i.e., the angle between the vertical velocity directions of the detection target and the lighting device). It is used to determine the up and down position of the detection target in front of the lighting device.

[0132] The method for calculating the moving speed of the detection target based on the moving direction of the detection target, the moving speed of the light-carrying device, the center position of the detection target, and the distance between the detection target and the light-carrying device includes:

[0133] ;

[0134] ;

[0135] ;

[0136] ;

[0137] in, To detect the target moving speed, To detect the velocity component of the target on the X-axis, is the speed component of the light-carrying equipment on the X-axis, To detect the rate of change of the distance between the target and the light-carrying equipment, Indicates the time interval, To detect the target's velocity component on the Y axis, is the velocity component of the light-carrying equipment moving speed on the Y axis, To detect the target's velocity component on the Z axis, It is the velocity component of the light-carrying device on the Z axis.

[0138] The beam adjustment module is used to adaptively adjust the spotlight beam according to the parameter information of the lamp-borne equipment, the detection target category, the detection target feature data, the detection target movement speed and the detection target movement direction.

[0139] The method for adaptively adjusting the beam of a floodlight according to the parameter information of the onboard device, the category of the detection target, the characteristic data of the detection target, the moving speed of the detection target, and the moving direction of the detection target includes:

[0140] If the detection target category is a pedestrian, the spotlight beam is adaptively adjusted according to the detection target feature data, detection target movement speed and detection target movement direction;

[0141] If the detection target category is a lamp-mounted device, the spotlight beam is adaptively adjusted according to the detection target lamp-mounted device based on the detection target feature data, the detection target movement speed and the detection target movement direction.

[0142] like Figure 3 As shown, the method for adaptively adjusting the light beam of a floodlight for pedestrians includes:

[0143] Dynamically adjust the horizontal and vertical angle ranges of the light-mounted device based on the horizontal and vertical angle ranges of the detection target, so that the horizontal and vertical angle ranges of the light-mounted device cover the horizontal and vertical angle ranges of the detection target. If the original angle range already covers the detection target angle range, maintain the original setting without adjustment. The details are as follows:

[0144] If the lower limit of the horizontal angle range of the lamp-mounted device is less than or equal to the lower limit of the horizontal angle range of the detection target, and the upper limit of the horizontal angle range of the lamp-mounted device is greater than or equal to the upper limit of the horizontal angle range of the detection target, then the horizontal angle range of the lamp-mounted device does not need to be adjusted;

[0145] If the lower limit of the vertical angle range of the lamp-mounted device is less than or equal to the lower limit of the vertical angle range of the detection target, and the upper limit of the vertical angle range of the lamp-mounted device is greater than or equal to the upper limit of the vertical angle range of the detection target, then the vertical angle range of the lamp-mounted device does not need to be adjusted;

[0146] If neither the horizontal angle range nor the vertical angle range of the lamp-mounted device needs to be adjusted, the current process ends;

[0147] If the lower limit of the horizontal angle range of the lamp-mounted device is greater than the lower limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the lower limit of the horizontal angle range of the lamp-mounted device is the lower limit of the horizontal angle range of the detection target; if the upper limit of the horizontal angle range of the lamp-mounted device is less than the upper limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the upper limit of the horizontal angle range of the lamp-mounted device is the upper limit of the horizontal angle range of the detection target;

[0148] If the lower limit of the vertical angle range of the lamp-mounted equipment is greater than the lower limit of the vertical angle range of the detection target, the spotlight lens is adjusted so that the lower limit of the vertical angle range of the lamp-mounted equipment is the lower limit of the vertical angle range of the detection target; if the upper limit of the vertical angle range of the lamp-mounted equipment is less than the upper limit of the vertical angle range of the detection target, the spotlight lens is adjusted so that the upper limit of the vertical angle range of the lamp-mounted equipment is the upper limit of the vertical angle range of the detection target.

[0149] like Figure 4 As shown, the method for adaptively adjusting the light beam of a floodlight for detecting a target lamp-mounted device includes:

[0150] Dynamically adjust the horizontal and vertical angle ranges of the light-borne equipment based on the horizontal and vertical angle ranges of the detection target, so that the horizontal angle range of the light-borne equipment covers the horizontal angle range of the detection target, and the lower limit of the vertical angle range of the light-borne equipment is not higher than the lower limit of the vertical angle range of the detection target;

[0151] Input the detection target feature data, detection target movement speed, and detection target movement direction into a pre-built angle setting model to obtain the upper limit of the vertical angle prediction of the lamp-mounted device, and dynamically adjust the upper limit of the vertical angle range of the lamp-mounted device based on the upper limit of the vertical angle prediction; the details are as follows:

[0152] If the lower limit of the horizontal angle range of the lamp-mounted device is less than or equal to the lower limit of the horizontal angle range of the detection target, and the upper limit of the horizontal angle range of the lamp-mounted device is greater than or equal to the upper limit of the horizontal angle range of the detection target, then the horizontal angle range of the lamp-mounted device does not need to be adjusted;

[0153] If the lower limit of the horizontal angle range of the lamp-mounted device is greater than the lower limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the lower limit of the horizontal angle range of the lamp-mounted device is the lower limit of the horizontal angle range of the detection target; if the upper limit of the horizontal angle range of the lamp-mounted device is less than the upper limit of the horizontal angle range of the detection target, the lens of the spotlight is adjusted so that the upper limit of the horizontal angle range of the lamp-mounted device is the upper limit of the horizontal angle range of the detection target;

[0154] If the lower limit of the vertical angle range of the lamp-mounted device is less than or equal to the lower limit of the vertical angle range of the detection target, the lower limit of the vertical angle range of the lamp-mounted device does not need to be adjusted; if the lower limit of the vertical angle range of the lamp-mounted device is greater than the lower limit of the vertical angle range of the detection target, the lens of the floodlight is adjusted so that the lower limit of the vertical angle range of the lamp-mounted device is the lower limit of the vertical angle range of the detection target;

[0155] The detection target feature data, detection target movement speed and detection target movement direction are input into the pre-built angle setting model to obtain the predicted upper limit of the vertical angle of the lamp-mounted device; if the upper limit of the vertical angle range of the lamp-mounted device is equal to the predicted upper limit of the vertical angle of the lamp-mounted device, there is no need to adjust the upper limit of the vertical angle range of the lamp-mounted device; if the upper limit of the vertical angle range of the lamp-mounted device is not equal to the predicted upper limit of the vertical angle of the lamp-mounted device, the spotlight lens is adjusted so that the upper limit of the vertical angle range of the lamp-mounted device is the predicted upper limit of the vertical angle of the lamp-mounted device.

[0156] It's important to note that the lens, a key optical component in a floodlight, precisely controls the beam's illumination range by changing the path of the light beam emitted by the light source, thereby optimizing the lighting effect. Lens adjustment allows floodlights to flexibly adapt to various environments, providing the desired lighting effects to meet different usage requirements. Specifically, by adjusting the lens's horizontal rotation angle, the beam can be diverged or focused horizontally, thereby adjusting the beam's illumination range. By changing the lens's vertical rotation angle, the vertical distribution of the beam can be controlled.

[0157] Modern floodlights are equipped with adjustable optical components, such as liquid crystal optical switches (LCOS) and adjustable irises. These components can adjust the illumination range of the floodlight beam in real time according to control signals, thereby accurately matching the lighting needs of different targets, greatly improving the adaptability and accuracy of floodlights in complex environments.

[0158] The training method of the angle setting model includes:

[0159] Preliminarily collecting an angle setting data set, the angle setting data set comprising R groups of angle setting data and upper limits of vertical angle prediction corresponding to the R groups of angle setting data, where R is a positive integer greater than 0, the angle setting data comprising detection target feature data, detection target movement speed, and detection target movement direction; dividing the angle setting data set into a training set and a validation set, wherein the training set is used to train an angle setting model, and the validation set is used to evaluate the generalization performance of the angle setting model;

[0160] During the training process of the angle setting model, the optimization objective is to minimize the cross entropy loss function. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, training is stopped. The angle setting model is trained based on the logistic regression model or the support vector machine model.

[0161] The angle setting data is converted into a high-dimensional feature vector; the input layer of the angle setting model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through several hidden layers. Finally, the output layer of the angle setting model calculates the probability distribution of the upper limit of the vertical angle prediction through the softmax activation function, and outputs the upper limit of the vertical angle prediction corresponding to the maximum probability as the final prediction result.

[0162] It's important to note that the dynamic upper limit of vertical angle prediction is crucial during the operation of a headlight, as it directly impacts the driver's visual safety and driving experience. Because the road, target location, speed, and direction are all constantly changing, the upper limit of vertical angle prediction must be dynamically adjusted to effectively prevent direct light from reaching the driver's eyes.

[0163] Example 2

[0164] See also Figure 2 As shown, this embodiment provides a method for optimizing the illumination range of a floodlight lens, including:

[0165] Constructing the environment where the light-carrying device is located into a three-dimensional environment space, wherein the light-carrying device is a device having a floodlight lens;

[0166] Real-time collection of detection target parameter information;

[0167] According to the detection target parameter information, the detection target category is obtained;

[0168] Extracting features of the detection target parameter information according to the detection target category to obtain detection target feature data;

[0169] Real-time collection of lamp-borne equipment parameter information;

[0170] Processing the detection target parameter information and the light-borne equipment parameter information according to the detection target feature data to obtain the detection target moving speed and detection target moving direction;

[0171] The floodlight beam is adaptively adjusted according to the parameter information of the lamp-mounted equipment, the detection target category, the detection target feature data, the detection target movement speed and the detection target movement direction.

Claims

1. A method for optimizing the illumination range of a floodlight lens, characterized in that: include: Constructing the environment where the light-carrying device is located into a three-dimensional environment space, wherein the light-carrying device is a device having a floodlight lens; Real-time collection of detection target parameter information; According to the detection target parameter information, the detection target category is obtained; Extracting features from the detection target parameter information according to the detection target category to obtain detection target feature data; The center position of the detection target, the horizontal angle range of the detection target and the vertical angle range of the detection target are constructed into detection target feature data; Real-time collection of lamp-borne equipment parameter information; The parameter information of the lamp-mounted device includes the moving speed of the lamp-mounted device, the moving direction of the lamp-mounted device, the horizontal angle range of the lamp-mounted device, and the vertical angle range of the lamp-mounted device; Processing the detection target parameter information and the light-borne equipment parameter information according to the detection target feature data to obtain the detection target moving speed and detection target moving direction; Adaptively adjust the floodlight beam according to the parameter information of the onboard equipment, the detection target category, the detection target characteristic data, the detection target moving speed and the detection target moving direction; the adaptive adjustment of the floodlight beam includes adaptive adjustment according to the onboard equipment of the detection target; The method for adaptively adjusting the equipment on the target lamp is as follows: Dynamically adjust the horizontal and vertical angle ranges of the light-borne equipment based on the horizontal and vertical angle ranges of the detection target, so that the horizontal angle range of the light-borne equipment covers the horizontal angle range of the detection target, and the lower limit of the vertical angle range of the light-borne equipment is not higher than the lower limit of the vertical angle range of the detection target; The detection target feature data, detection target movement speed and detection target movement direction are input into a pre-built angle setting model to obtain the vertical angle prediction upper limit of the lamp-borne equipment, and the upper limit of the vertical angle range of the lamp-borne equipment is dynamically adjusted according to the vertical angle prediction upper limit.

2. The method for optimizing the illumination range of a floodlight lens according to claim 1, wherein: The method for adaptively adjusting the beam of a floodlight according to the parameter information of the onboard device, the category of the detection target, the characteristic data of the detection target, the moving speed of the detection target, and the moving direction of the detection target includes: If the detection target category is a pedestrian, the spotlight beam is adaptively adjusted according to the detection target feature data, detection target movement speed and detection target movement direction; If the detection target category is a detection target lamp-mounted device, the spotlight beam is adaptively adjusted for the detection target lamp-mounted device based on the detection target feature data, the detection target moving speed, and the detection target moving direction.

3. The method for optimizing the illumination range of a floodlight lens according to claim 2, wherein: Methods for adaptively adjusting the floodlight beam for pedestrians include: According to the horizontal angle range and vertical angle range of the detection target, the horizontal angle range and vertical angle range of the lamp-mounted equipment are dynamically adjusted so that the horizontal angle range and vertical angle range of the lamp-mounted equipment cover the horizontal angle range and vertical angle range of the detection target; if the original angle range already covers the angle range of the detection target, the original setting is maintained without adjustment.

4. The method for optimizing the illumination range of a floodlight lens according to claim 3, wherein: The method for acquiring the detection target moving speed and the detection target moving direction includes: Obtain the center position of the detection target from the detection target feature data, obtain the three-dimensional coordinates of the light-carrying device from the three-dimensional environment space; obtain the distance between the detection target and the light-carrying device from the detection target parameter information; The horizontal speed direction and the vertical speed direction of the detection target are calculated based on the center position of the detection target and the three-dimensional coordinates of the light-carrying device; the horizontal speed direction and the vertical speed direction of the detection target are constructed into the movement direction of the detection target; Obtain the moving speed of the lamp-borne equipment from the lamp-borne equipment parameter information; The moving speed of the detection target is calculated based on the moving direction of the detection target, the moving speed of the lamp-mounted device, the center position of the detection target, and the distance between the detection target and the lamp-mounted device.

5. The method for optimizing the illumination range of a floodlight lens according to claim 4, wherein: The method for acquiring the detection target feature data includes: Analyze and process the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height; The center position of the detection target is calculated according to the target refinement starting position, the detection target refinement length, the detection target refinement width and the detection target refinement height; The horizontal angle range and vertical angle range of the detection target are calculated according to the center position of the detection target, the refined width of the detection target and the refined height of the detection target; The center position of the detection target, the horizontal angle range of the detection target and the vertical angle range of the detection target are constructed into the detection target feature data.

6. The method for optimizing the illumination range of a floodlight lens according to claim 5, characterized in that: The method of analyzing and processing the detection target category and detection target parameter information according to a preset method to obtain the target refinement starting position, detection target refinement length, detection target refinement width, and detection target refinement height includes: Extract the starting position and spatial size based on the three-dimensional coordinate data of the detection target; If the detection target is a detection target lamp-mounted device, based on the reflection intensity of the detection target, data points whose reflection intensity is within a preset glass reflection intensity range are extracted to form a glass area, and based on the glass area, the starting position and spatial size are extracted as the refined features of the detection target, and the target refined starting position, detection target refined length, detection target refined width and detection target refined height are obtained.

7. The method for optimizing the illumination range of a floodlight lens according to claim 6, wherein: The method for acquiring the detection target category includes: Inputting the detection target parameter information into a pre-built category diagnosis model to obtain the detection target category, wherein the detection target category includes pedestrians and detection target light-mounted equipment; The training method of the category diagnosis model includes: Pre-collecting a category diagnosis dataset, the category diagnosis dataset including Q groups of category diagnosis data and detection target categories corresponding to the Q groups of category diagnosis data, where Q is a positive integer greater than 0, and the category diagnosis data including detection target parameter information; dividing the category diagnosis dataset into a training set and a validation set, wherein the training set is used to train a category diagnosis model, and the validation set is used to evaluate the generalization performance of the category diagnosis model; During the training process of the category diagnosis model, minimizing the cross-entropy loss function is used as the optimization goal. An early stopping strategy is used to monitor the performance of the validation set, and the network parameters are continuously adjusted to optimize the model performance. When the prediction accuracy on the validation set reaches the expected accuracy, training is stopped. The category diagnosis model is trained based on a logistic regression model or a support vector machine model. The category diagnosis data is converted into a high-dimensional feature vector; the input layer of the category diagnosis model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through the hidden layer. Finally, the output layer of the category diagnosis model calculates the probability distribution of the detection target category through the softmax activation function, and outputs the detection target category corresponding to the maximum probability as the final prediction result.

8. The method for optimizing the illumination range of a floodlight lens according to claim 7, wherein: The method for constructing the three-dimensional environment space includes: The radar sensor on the light-mounted device scans the environment, collects the position and shape characteristics of objects in the environment, and generates raw point cloud data. The raw point cloud data consists of discrete data points, each of which corresponds to a location in the environment and records the three-dimensional coordinate information of the data point. The collected original point cloud data is cleaned of noise, invalid data in the original point cloud data is removed, and valid data point information in the original point cloud data is retained to obtain the preprocessed environmental point cloud data; Convert the pre-processed environmental point cloud data into an occupancy grid model, set the grid unit size for the occupancy grid model, divide the occupancy grid model into regular grid units with equal spacing according to the grid unit size; assign data points to corresponding regular grid units to complete the conversion from point data to unit data; The regular grid cells occupied by obstacles are marked, and the marked regular grid cells are integrated into a three-dimensional map structure processed by a computer to obtain a three-dimensional environment space.

9. The method for optimizing the illumination range of a floodlight lens according to claim 8, wherein: The training method of the angle setting model includes: Preliminarily collecting an angle setting data set, the angle setting data set comprising R groups of angle setting data and upper limits of vertical angle prediction corresponding to the R groups of angle setting data, where R is a positive integer greater than 0, the angle setting data comprising detection target feature data, detection target movement speed, and detection target movement direction; dividing the angle setting data set into a training set and a validation set, wherein the training set is used to train an angle setting model, and the validation set is used to evaluate the generalization performance of the angle setting model; During the training process of the angle setting model, the optimization objective is to minimize the cross entropy loss function. An early stopping strategy is used to monitor the performance of the validation set, and the model performance is optimized by continuously adjusting the network parameters. Training is stopped when the prediction accuracy on the validation set reaches the expected accuracy. The angle setting model is trained based on the logistic regression model or the support vector machine model. The angle setting data is converted into a high-dimensional feature vector; the input layer of the angle setting model receives the high-dimensional feature vector, and the nonlinear relationship in the data is extracted through the hidden layer. Finally, the output layer of the angle setting model calculates the probability distribution of the upper limit of the vertical angle prediction through the softmax activation function, and outputs the upper limit of the vertical angle prediction corresponding to the maximum probability as the final prediction result.

10. Floodlight lens, characterized in that: include: Based on the method for optimizing the lighting range of the floodlight lens as described in any one of claims 1 to 9, adaptive adjustment of the floodlight beam is achieved according to the parameter information of the lamp-borne equipment, the detection target category, the detection target feature data, the detection target movement speed and the detection target movement direction.

Images