CN119018043B - Adaptive headlight control method and adaptive headlight system - Google Patents
Adaptive headlight control method and adaptive headlight system Download PDFInfo
- Publication number
- CN119018043B CN119018043B CN202411072318.0A CN202411072318A CN119018043B CN 119018043 B CN119018043 B CN 119018043B CN 202411072318 A CN202411072318 A CN 202411072318A CN 119018043 B CN119018043 B CN 119018043B
- Authority
- CN
- China
- Prior art keywords
- distance
- target
- headlamp
- confidence
- optimal angle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60Q—ARRANGEMENT OF SIGNALLING OR LIGHTING DEVICES, THE MOUNTING OR SUPPORTING THEREOF OR CIRCUITS THEREFOR, FOR VEHICLES IN GENERAL
- B60Q1/00—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor
- B60Q1/02—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments
- B60Q1/04—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights
- B60Q1/06—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights adjustable, e.g. remotely-controlled from inside vehicle
- B60Q1/08—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights adjustable, e.g. remotely-controlled from inside vehicle automatically
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60Q—ARRANGEMENT OF SIGNALLING OR LIGHTING DEVICES, THE MOUNTING OR SUPPORTING THEREOF OR CIRCUITS THEREFOR, FOR VEHICLES IN GENERAL
- B60Q1/00—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor
- B60Q1/02—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments
- B60Q1/04—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights
- B60Q1/14—Arrangement of optical signalling or lighting devices, the mounting or supporting thereof or circuits therefor the devices being primarily intended to illuminate the way ahead or to illuminate other areas of way or environments the devices being headlights having dimming means
- B60Q1/1415—Dimming circuits
- B60Q1/1423—Automatic dimming circuits, i.e. switching between high beam and low beam due to change of ambient light or light level in road traffic
- B60Q1/143—Automatic dimming circuits, i.e. switching between high beam and low beam due to change of ambient light or light level in road traffic combined with another condition, e.g. using vehicle recognition from camera images or activation of wipers
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Databases & Information Systems (AREA)
- Mechanical Engineering (AREA)
- Multimedia (AREA)
- Medical Informatics (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Lighting Device Outwards From Vehicle And Optical Signal (AREA)
Abstract
The invention relates to a control method of a self-adaptive headlamp. The control method of the self-adaptive headlamp comprises the steps of obtaining a target distance in real time, carrying out mapping relation conversion on the target distance by adopting an optimal angle control model to obtain a corresponding headlamp angle and a corresponding lamp gear, and controlling the headlamp according to the headlamp angle and the lamp gear control. The optimal angle control model is a fitting function of the distance mapping headlamp angles, and different lamplight gears are output according to the distances and the headlamp angles. According to the control method of the self-adaptive headlamp, the corresponding lamplight gear and the optimal headlamp angle are directly mapped through the target distance acquired in real time and the optimal angle control model, so that the headlamp is adjusted in real time, the environment change is responded rapidly, and the dynamic adaptability of the system is improved. In addition, the invention can be easily deployed to an embedded system and has high real-time performance.
Description
Technical Field
The present invention relates to the field of automotive adaptive headlamps, and more particularly, to a control method of an adaptive headlamp, an adaptive headlamp system, an electronic device, and a computer storage medium.
Background
The far and near light switching and the angle adjustment of the traditional car lamp are manually adjusted after the driver observes the surrounding environment. However, in a low-light environment, particularly in a situation where the driving speed is high, when the driver focuses on the environment, there is a certain safety hazard through the operations of manually switching the high beam and the low beam and adjusting the angle of the lamp. In addition, the manually adjusted high beam is not used properly, so that glare to a driver, a pedestrian or a rider of the vehicle is easily caused, and a blind area of a sight line can be generated due to improper adjustment of the angle of the lamp, so that the running risk of the driver is increased.
In order to solve the potential safety hazard or error caused by manual adjustment of the adjustment automotive headlamp, a deep learning algorithm is generally combined with an automotive headlamp control system in the prior art to realize an Adaptive Front-LIGHTING SYSTEM (AFS) of an automotive vehicle so as to solve the problem that the adjustment automotive headlamp is required to be manually adjusted. However, in practical application, a multi-mode deep learning algorithm or a high-precision deep learning algorithm is adopted, so that the method generally has the defects of high computational complexity and high computational complexity, difficult deployment of an embedded system, reduced instantaneity and the like.
Disclosure of Invention
Based on this, an object of the present invention is to provide a control method of an adaptive front light.
A control method of an adaptive front lighting lamp, comprising the steps of:
S1, acquiring a target distance in real time;
s2, performing mapping relation conversion on the target distance by adopting an optimal angle control model to obtain a corresponding headlamp angle and a lamp gear, and controlling the headlamp according to the headlamp angle and the lamp gear, wherein the optimal angle control model is a fitting function of distance-mapping headlamp angle, and outputs different lamp gears according to the distance and the headlamp angle, and specifically comprises the following steps:
Being equipped with n light gears, then:
f(x)=θ
light=αn,if x∈[dn-1,dn]and θn-1≤θ≤θn
Wherein f (x) is an optimal angle control model of an input target distance x, θ is an output optimal angle corresponding to the input x, light=α n is an nth light gear, d n-1,dn is a target distance range used in the light gear α n, and θ n-1≤θ≤θn is a headlamp angle range output in the optimal angle control model in the light gear α n.
The control method of the self-adaptive headlamp provided by the invention has the advantages that the headlamp is adjusted in real time by directly mapping the corresponding lamplight gear and the optimal headlamp angle through the optimal angle control model, the dynamic adaptability of the system is improved by quickly responding to environmental changes, and meanwhile, the optimal angle control model provided by the invention has good expansibility by using a function model based on mathematical modeling, can be expanded according to different sensors and data sources, and constructs a multi-mode control model, thereby adapting to new environments and demands and improving the control flexibility and adaptability of the self-adaptive headlamp.
Further, the construction steps of the optimal angle control model are as follows:
T1, acquiring video data of headlamps with different illumination intensities and the same travel mode;
the video data are front view image pictures of a test vehicle driving path acquired by shooting equipment;
t2, performing target detection on each frame of video data to obtain confidence coefficient data of a plurality of correct detection results;
T3, carrying out data cleaning on confidence data of a plurality of correct detection results to obtain a plurality of cleaned confidence data;
T4, performing function fitting on the cleaned confidence data to obtain a series of continuous curves, and converting the frame number into a distance according to the same travel mode of the test vehicle to obtain a mapping relation between the distance and the confidence;
T5, screening the mapping relation between the distance and the confidence coefficient according to a confidence coefficient threshold value to obtain a plurality of optimal angle key points;
And T6, performing function fitting on a plurality of optimal angle key points to obtain an optimal angle control model.
According to the invention, the influence of headlamps with different illumination intensities on the target detection algorithm is evaluated by the confidence coefficient of the target detection algorithm to carry out mathematical modeling, and meanwhile, the mathematical model is ensured to be applied and deployed under the high confidence coefficient, so that the reliability and the accuracy of the mathematical model are ensured.
Further, the step T5 includes the following substeps:
T51, setting a confidence threshold value Cof;
t52, screening points with confidence coefficient greater than or equal to Cof in the mapping relation between the distance and the confidence coefficient, and taking the points as a plurality of candidate key points;
And T53, screening out the lowest angle in the candidate key points to serve as a plurality of optimal angle key points, wherein the optimal angle key points meet the following conditions:
θ=min{θ|c(d,θ,α)≥Cof}
where c (d, θ, α) is used to represent the target detection confidence function at distance d, longitudinal illumination angle θ, and light gear α.
According to the invention, the point with the confidence coefficient (ordinate) greater than or equal to Cof in the relation curve of the distance and the confidence coefficient is selected as the candidate key point, and the key point corresponding to the minimum angle is selected, so that the head lamp can provide enough detection confidence coefficient under different distances, and meanwhile, the influence on glare of other people is reduced as much as possible.
Further, the target distance is obtained through an imaging device and a lightweight target detection algorithm, and the method specifically comprises the following steps:
SA11, recognizing an image through a lightweight target detection algorithm to obtain pixel coordinate information of a target;
SA12, positioning the target through pixel coordinate information of the target to obtain distance information of the target;
and SA13, carrying out distance measurement and data compensation by adopting a sensor according to the distance information of the target to obtain the target distance.
The invention obtains the accurate distance between the target and the target by combining a lightweight target detection algorithm, so as to further improve the intellectualization of the self-adaptive headlamp, and ensures the accuracy of the target distance by reducing the calculated amount of the target detection algorithm and by sensor ranging compensation, thereby fully reducing the implementation cost.
Further, the target distance is obtained through a distance measuring sensor, so that the calculation cost is reduced.
An adaptive headlamp system comprises a headlamp, a motor, a lamplight gear adjusting mechanism, a ranging module and a controller electrically and/or communicatively connected with the motor, the lamplight gear adjusting mechanism and the ranging module;
the motor is used for controlling the longitudinal illumination angle of the head lamp according to the angle of the head lamp;
the lamplight gear adjusting mechanism is used for controlling the lamplight gear of the headlamp according to the lamplight gear;
The controller is used for acquiring a target distance in real time through the ranging module, and converting a mapping relation of the target distance by adopting an optimal angle control model to obtain a corresponding headlamp angle and a lamp gear, wherein the optimal angle control model is a fitting function of the distance-mapping headlamp angle, and outputs different lamp gears according to the distance and the headlamp angle, and specifically comprises the following steps:
Being equipped with n light gears, then:
f(x0=θ
light=αn,if x∈[dn-1,dn]and θn-1≤θ≤θn
Wherein f (x) is an optimal angle control model of an input target distance x, θ is an output optimal angle corresponding to the input x, light=α n is an nth light gear, d n-1,dn is a target distance range used in the light gear α n, and θ n-1≤θ≤θn is a headlamp angle range output in the optimal angle control model in the light gear α n.
Further, the obtaining step of the optimal angle control model is as follows:
T1, acquiring video data of headlamps with different illumination intensities and the same travel mode;
the video data are front view image pictures of a test vehicle driving path acquired by shooting equipment;
t2, performing target detection on each frame of video data to obtain confidence coefficient data of a plurality of correct detection results;
T3, carrying out data cleaning on confidence data of a plurality of correct detection results to obtain a plurality of cleaned confidence data;
T4, performing function fitting on the cleaned confidence data to obtain a series of continuous curves, and converting the frame number into a distance according to the same travel mode of the test vehicle to obtain a mapping relation between the distance and the confidence;
T5, screening the mapping relation between the distance and the confidence coefficient according to a confidence coefficient threshold value to obtain a plurality of optimal angle key points;
And T6, performing function fitting on a plurality of optimal angle key points to obtain an optimal angle control model.
Further, the step T5 includes the following substeps:
T51, setting a confidence threshold value Cof;
t52, screening points with confidence coefficient greater than or equal to Cof in the mapping relation between the distance and the confidence coefficient, and taking the points as a plurality of candidate key points;
And T53, screening out the lowest angle in the candidate key points to serve as a plurality of optimal angle key points, wherein the optimal angle key points meet the following conditions:
θ=min{θ|c(d,θ,α)≥Cof}
where c (d, θ, α) is used to represent the target detection confidence function at distance d, longitudinal illumination angle θ, and light gear α.
Further, the ranging module is an image capturing device and a lightweight target detection algorithm, and obtains the target distance through the following steps:
SA11, recognizing an image through a lightweight target detection algorithm to obtain pixel coordinate information of a target;
SA12, positioning the target through pixel coordinate information of the target to obtain distance information of the target;
And SA13, according to the distance information of the target, adopting a sensor to perform distance measurement and data compensation, and obtaining the target distance.
Further, the ranging module is a ranging sensor.
For a better understanding and implementation, the present invention is described in detail below with reference to the drawings.
Drawings
FIG. 1 is a simplified schematic diagram of an adaptive front lighting system according to the present invention;
FIG. 2 is a simplified flow chart of a control method of the adaptive front lighting lamp according to FIG. 1;
FIG. 3 is a flow chart of the construction of the optimal angle control model according to the present invention;
FIG. 4 is a schematic diagram of a simulated example of a multi-distance-confidence curve;
FIG. 5 is a schematic diagram of an example of a curve trend simulation based on the optimal angle control model of FIG. 4;
fig. 6 is a schematic diagram showing the simple implementation effect of the adaptive front lighting system according to the present invention.
Detailed Description
In order to solve the problems of high complexity, difficult deployment of an embedded system and low instantaneity of a high-precision deep learning algorithm in the prior art, the invention provides a self-adaptive headlamp system, which comprises an improved method for controlling the gear and the angle of a headlamp, the control method adopts an optimal angle control model fitted by a function, the obtained target distance is converted in real time to obtain the optimal angle and the lamplight gear corresponding to the target distance, so that the algorithm complexity in the prior art is reduced, the embedded system is simple to deploy, and the real-time property of the self-adaptive headlamp adjustment is improved on the basis of realizing the self-adaptive headlamp adjustment of the automobile.
Referring to fig. 1, the adaptive headlamp system of the present invention includes a headlamp, a motor, a light gear adjusting mechanism, a ranging module, and a controller electrically and/or communicatively connected to the motor, the light gear adjusting mechanism, and the ranging module.
The motor can be a stepping motor, and the longitudinal illumination angle of the headlamp is controlled according to a given corresponding optimal headlamp angle (expected deflection angle), so that the precise control of the headlamp angle is realized.
The light gear adjusting mechanism can be selected to be wave width control dimming (Pulse Width Modulation, PWM), and controls the light gear of the headlamp according to a given light gear, and particularly changes the average light current of the lamp by adjusting the pulse width, so that the illumination intensity of the headlamp is adjusted. However, since the control method of the headlight corresponds to different scene requirements (quick response, accuracy or cost control, etc.), the selection of the control means is different, and thus the present invention is not limited to the control means specifically.
In an embodiment, the distance measuring module is a distance measuring sensor, which is used for measuring the distance between the object and the object, and the distance measuring module is used for selecting laser distance measurement, laser point cloud, ultrasonic sensor and the like, or obtaining the object distance with higher precision through combination of multiple sensors, so that the invention is not particularly limited to the selection of the sensor.
In another embodiment, the distance measuring module is an image capturing device and a lightweight target detection algorithm, and the target distance is obtained by processing a target image captured by the image capturing device in real time through the lightweight target detection algorithm.
Referring to fig. 2, the controller obtains the gear of the light corresponding to the target and the optimal headlamp angle through the following steps.
S1, acquiring a target distance in real time through the ranging module.
In an embodiment, when the ranging module is used as the ranging module for the camera device and the lightweight object detection algorithm, the steps for obtaining the object distance are as follows:
and SA11, identifying the image through a lightweight object detection algorithm to obtain the pixel coordinate information of the object.
The lightweight object detection algorithm may optionally be YOLOv (You only look Once Version 8), identify an object in the current frame, and determine a bounding box for the object to obtain pixel coordinate information for the object. The YOLOv target detection algorithm comprises a feature extraction module (Cross STAGE PARTIAL Network, CSPNet), a feature fusion module (Path Aggregation Network, PANet) and a detection module, which are respectively used for extracting features, fusing features and detecting and judging, and outputting target types, target bounding boxes and corresponding confidence data.
The invention takes pedestrians, traffic signs and vehicles as the primary detection targets for detection, but the primary detection targets can be further increased or changed according to different environments and requirements, for example, the primary detection targets are applied to autonomous mobile devices, unmanned aerial vehicles and the like, and the preset target types of users are detected. Therefore, the present invention is not particularly limited herein as to the type of target it detects.
SA12, positioning the target through the pixel coordinate information of the target to obtain the distance information of the target.
The positioning can be selected as monocular ranging, a boundary box of the target is identified according to a target detection algorithm, and the distance corresponding to the size of the boundary box of the current target is estimated by combining calibration data (the size and the focal length of the corresponding shooting equipment), so that the world coordinate position of the target is estimated and is used as the target distance.
Further, the method also comprises a step SA13 of adopting a sensor to perform distance measurement and data compensation according to the distance information of the target to obtain the target distance so as to obtain the target distance with higher precision. The data compensation may be performed by summing the distance obtained by combining the ranging of the different sensors with the distance information of the current target and taking an average value.
And S2, carrying out mapping relation conversion on the input target distance by adopting an optimal angle control model to obtain a corresponding headlamp angle and a corresponding lamplight gear.
Specifically, the optimal angle control model is a fitting function of the distance mapping headlamp angles, and has a lamplight gear condition, and different lamplight gears are output according to the relationship between the distance and the headlamp angles, for example, if the lamplight gears have n lamplight gears, the following expression can be referred to:
f(x)=θ
light=αn,if x∈[dn-1,dn]and θn-1≤θ≤θn
The method comprises the steps of (a) setting a target distance range of a lamp, wherein f (x) is an optimal angle control model of an input target distance x, theta is an output optimal angle corresponding to the input x, alpha n is an nth lamp gear, (d n-1,dn) is a target distance range using a current lamp gear alpha n, a plurality of target distance ranges can be arranged according to different lamp angles, and theta n-1≤θ≤θn is a lamp angle range output in the optimal angle control model under the lamp gear alpha n.
The invention adopts the function model based on mathematical modeling to obtain the control parameters of the head lamp, so that the control of the self-adaptive head lamp can rapidly respond to the environmental change, thereby improving the driving safety and comfort. In addition, the function model based on mathematical modeling has good expansibility, can be expanded according to different sensors and data sources, and constructs a multi-mode control model so as to adapt to new environments and requirements, and improve the control flexibility and adaptability of the adaptive headlamp.
In order to ensure that the optimal headlamp angle and the light gear output by the optimal angle control model can effectively reduce the glare influence on other vehicles or pedestrians, and simultaneously ensure that a driver is in the maximum visible range in a low-light environment, the construction steps of the optimal angle control model are as follows:
and T1, acquiring video data of headlamps with different illumination intensities and the same travel mode.
Specifically, the video data is a front image picture of a travel path of a test vehicle acquired by a photographing apparatus, the test vehicle starting from a fixed starting point and traveling in the same route and travel mode (such as speed, acceleration, etc.). The illumination intensity is the combination of different illumination angles and light gears of the test head lamps.
In order to ensure consistency and comparability of the collected video data, the invention adjusts the illumination intensity as a unique variable, and all video data have the same resolution and frame rate so as to evaluate the influence of different headlamps on the visibility of a driver.
Besides taking illumination intensity as a unique variable, the method can further collect video data through multiple weather conditions or collect video data through dynamic illumination simulation so as to simulate the influence of more variables on the illumination intensity of the test headlamp.
And T2, carrying out target detection on each frame of video data to obtain confidence coefficient data of a plurality of correct detection results.
Specifically, video data are grouped according to illumination intensity, and synchronous framing is performed to obtain a plurality of groups of video frames, wherein the groups of video frames are used for keeping the consistency of time and space between the frames.
And carrying out target detection on a plurality of groups of video frames to identify target categories in the frames, such as vehicles, pedestrians or traffic signs, and the like, and filtering out the target categories with the identification errors to obtain confidence coefficient data of the correct detection results of a plurality of groups.
The confidence data is a numerical value output by the target detection model and is used for representing the similarity degree of the detection category and the actual category of the target in the current frame, so that the visual degree of the frame is represented.
Wherein the target detection is optionally performed based on a deep learning model, such as YOLOv (You Only Look Once version 8). The model comprises a feature extraction module, a feature fusion module and a detection module which are respectively used for extracting features, fusing features and detecting and judging, and outputting target types and confidence data thereof. The invention evaluates the influence of different illumination intensities on target detection by adopting confidence data. However, the present invention does not limit the specific model, since the target detection model is selected in various ways and can be adjusted according to the requirements (different accuracies).
And T3, carrying out data cleaning on the confidence data of the correct detection results to obtain cleaned confidence data.
Specifically, by processing confidence data of the correct detection results of a plurality of groups, a plurality of groups of confidence scatter diagrams are obtained. The abscissa of the scatter diagram is the number of frames, and the ordinate is the confidence data of the correct detection result.
And then, carrying out abnormal point detection on the plurality of groups of confidence scatter diagrams to obtain abnormal data.
And carrying out data cleaning on the plurality of groups of confidence scatter diagrams according to the abnormal data to obtain a plurality of cleaned confidence data.
The outlier detection may be based on algorithms such as distance, similarity, probability or cluster, for example, local outlier factor (Local Outlier Factor, LOF), isolated Forest (Isolation Forest), clustering algorithm (K-means), or proximity algorithm (K nearest neighbor), and the like, and therefore, the invention is not limited to the means of outlier detection specifically herein.
Because of accidental factors or environmental interference, points which deviate from the overall data distribution rule may appear in the confidence scatter diagram, and the abnormal points need to be removed through a data cleaning process so as to ensure that the data reflect the general rule in the general scene.
And T4, performing function fitting on the plurality of cleaned confidence data to obtain a mapping relation between the distance and the confidence.
Specifically, a series of continuous curves is obtained by functionally fitting the cleaned confidence data. The curves are used to represent the trend of confidence over the number of frames at different illumination intensities.
And then, converting the frame number into the distance according to the travel mode fixed by the test vehicle, so as to obtain a relation curve of the distance and the confidence coefficient, namely a mapping relation of the distance and the confidence coefficient.
The test vehicle runs at a constant speed and a fixed route, so that the distance corresponding to the target in each frame of image can be calculated according to the speed and the time. While the function fit may be selected as a polynomial fit or other fitting method to generate a continuous confidence curve, the means of function fit is not particularly limited herein.
And T5, screening the mapping relation between the distance and the confidence coefficient according to a confidence coefficient threshold value to obtain a plurality of optimal angle key points.
Specifically, a confidence threshold Cof is set for controlling the confidence of the detection result to be greater than Cof.
And in the mapping relation between the screening distance and the confidence coefficient, the point with the ordinate larger than or equal to Cof is used as a candidate key point.
And screening out the lowest angle in the candidate key points to serve as a plurality of optimal angle key points, wherein the expression can be referred to as follows:
θ=min{θ|c(d,θ,α)≥Cof}
where c (d, θ, α) is used to represent the target detection confidence function at distance d, longitudinal illumination angle θ, and light gear α.
The value of Cof can be selected to be 0.8 as the accuracy of the recognition closest to the human eye. However, according to the requirements of different road conditions, detection precision, safety level and the like, different confidence coefficient thresholds can be selected as the confidence coefficient of the control detection result to be always larger than the threshold, so that the invention is not particularly limited in the selection of the Cof value.
As each light gear and longitudinal illumination angle have a continuous distance-confidence relation curve, and the curves are in the same coordinate graph, a cluster of curves with the same trend is formed, and specific reference can be made to fig. 4, wherein fig. 4 is an example drawing through analog sampling data, and is not real data, but only for reference, in a certain distance interval, a plurality of curve confidence values are higher than a Cof threshold, and the light gear and longitudinal illumination angle corresponding to the curve with the lowest angle value are selected, so that the head lamp can provide enough detection confidence under different distances, and meanwhile, the influence on glare of other people is reduced as much as possible.
And T6, performing function fitting on the plurality of optimal angle key points to obtain an optimal angle control model.
Specifically, fitting and drawing are performed according to the optimal angle key points corresponding to the distances, so as to obtain a continuous distance-optimal angle curve, wherein the curve trend of the continuous distance-optimal angle curve is an optimal angle control model, and can be specifically referred to fig. 5.
The optimal angle control model is consistent with step S2, and the description of the present invention is not repeated here.
The invention can obtain the distance-correct target confidence coefficient curve through the target detection model, and obtain the optimal angle control model in a mathematical modeling mode, and can also use the multi-sensor data as the input of the multi-mode fusion target detection model, so that the distance-correct target confidence coefficient curve obtained through the fusion detection method can be expanded to serve as the data in the mathematical modeling process, and be used for replacing or supplementing the distance-optimal angle control model used by the invention, thereby improving the expandability and accuracy of the optimal angle control model.
Compared with the prior art, the method has the advantages that the target detection method is combined with mathematical modeling, so that the deployment process of the self-adaptive headlamp system is simplified, the accuracy and the reliability of data processing are improved, and the driving safety and the response speed of the self-adaptive headlamp system are ensured. In addition, the target detection models are isolated in a mathematical modeling mode, so that a user can select a multi-mode target detection model with high calculation amount according to different requirements, more complex scenes and requirements are met, and good expansion capacity is provided to cope with wider application scenes and requirements.
Meanwhile, the optimal angle control model is deployed to the self-adaptive headlamp system in an off-line mode, so that the vehicle can be ensured to be positioned in areas such as tunnels, suburbs and the like which lack signal amplification stations and base stations, and the self-adaptive headlamp system can be maintained to provide high-precision and high-real-time self-adaptive adjustment of headlamps.
In addition, along with the development of the target detection algorithm, a multi-mode fusion target detection method using multi-sensor data as input has become a mainstream gradually, and accordingly, a distance-correct target confidence curve obtained by fusing a plurality of different visual target detection methods can be easily expanded to serve as data in a modeling process and be used for replacing or supplementing a current optimal angle model, namely a distance-optimal angle curve, so that the expandability and the accuracy of a function model are fully improved.
Based on the same inventive concept, the present application also provides an electronic device, which may be a terminal device such as a server, a desktop computing device, or a mobile computing device (e.g., a laptop computing device, a handheld computing device, a tablet computer, a netbook, etc.). The device comprises one or more processors for executing programs to implement the method for controlling the adaptive headlamp according to the embodiment of the application, and a memory for storing a computer program executable by the processors.
Based on the same inventive concept, the present application also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the control method of an adaptive headlamp described in any of the above embodiments, corresponding to the embodiments of the control method of an adaptive headlamp described above.
The present application may take the form of a computer program product embodied on one or more storage media (including, but not limited to, magnetic disk storage, CD-ROM, optical storage, etc.) having program code embodied therein. Computer-usable storage media include both permanent and non-permanent, removable and non-removable media, and information storage may be implemented by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium which can be used to store information that can be accessed by a computing device.
The above examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that modifications and improvements can be made by those skilled in the art without departing from the spirit of the invention, and the invention is intended to encompass such modifications and improvements.
Claims (6)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411072318.0A CN119018043B (en) | 2024-08-06 | 2024-08-06 | Adaptive headlight control method and adaptive headlight system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411072318.0A CN119018043B (en) | 2024-08-06 | 2024-08-06 | Adaptive headlight control method and adaptive headlight system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119018043A CN119018043A (en) | 2024-11-26 |
| CN119018043B true CN119018043B (en) | 2025-02-25 |
Family
ID=93527820
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411072318.0A Active CN119018043B (en) | 2024-08-06 | 2024-08-06 | Adaptive headlight control method and adaptive headlight system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119018043B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119389103A (en) * | 2025-01-02 | 2025-02-07 | 常州星宇车灯股份有限公司 | Lamp stand integrated controller system and energy-saving control method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109849773A (en) * | 2018-12-29 | 2019-06-07 | 芜湖鑫芯微电子有限公司 | A kind of intelligent automotive light self-adapted adjustment system and method based on chip |
| CN115447478A (en) * | 2022-09-01 | 2022-12-09 | 浙江吉利控股集团有限公司 | Control method, control device, control system and vehicle |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE10334613A1 (en) * | 2003-07-29 | 2005-02-17 | Robert Bosch Gmbh | Driving auxiliary device |
| DE102009054227B4 (en) * | 2009-11-21 | 2025-10-02 | Volkswagen Ag | Method for controlling a headlight assembly for a vehicle and headlight assembly |
| CN201659924U (en) * | 2010-04-19 | 2010-12-01 | 吉林大学 | Adaptive system of headlight brightness |
| KR102639790B1 (en) * | 2019-03-28 | 2024-02-22 | 현대모비스 주식회사 | Head lamp control device and method |
| CN115534801B (en) * | 2022-08-29 | 2023-07-21 | 深圳市欧冶半导体有限公司 | Car lamp self-adaptive dimming method and device, intelligent terminal and storage medium |
| CN115601680A (en) * | 2022-10-28 | 2023-01-13 | 长沙海信智能系统研究院有限公司（Cn） | Optimal frame extraction method and device and electronic equipment |
| CN117445793A (en) * | 2023-10-16 | 2024-01-26 | 重庆赛力斯新能源汽车设计院有限公司 | Control method and device of high beam, electronic equipment and readable storage medium |
| CN117246226A (en) * | 2023-11-03 | 2023-12-19 | 奇瑞新能源汽车股份有限公司 | Anti-dazzle self-adaptive high beam headlamp control method and system |
| CN118061900A (en) * | 2024-03-30 | 2024-05-24 | 重庆赛力斯凤凰智创科技有限公司 | Control method and device for vehicle lamplight, electronic equipment and readable storage medium |
| CN118366130B (en) * | 2024-06-19 | 2024-10-01 | 深圳拜波赫技术有限公司 | Pedestrian glare protection and intelligent shadow area generation method and system |
-
2024
- 2024-08-06 CN CN202411072318.0A patent/CN119018043B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109849773A (en) * | 2018-12-29 | 2019-06-07 | 芜湖鑫芯微电子有限公司 | A kind of intelligent automotive light self-adapted adjustment system and method based on chip |
| CN115447478A (en) * | 2022-09-01 | 2022-12-09 | 浙江吉利控股集团有限公司 | Control method, control device, control system and vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119018043A (en) | 2024-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7369921B2 (en) | Object identification systems, arithmetic processing units, automobiles, vehicle lights, learning methods for classifiers | |
| CN108513059B (en) | Image processing method, device and automatic driving vehicle | |
| JP7371053B2 (en) | Electronic devices, mobile objects, imaging devices, and control methods, programs, and storage media for electronic devices | |
| TW202019745A (en) | Systems and methods for positioning vehicles under poor lighting conditions | |
| CN110045729A (en) | A kind of Vehicular automatic driving method and device | |
| US20250086954A1 (en) | Correction of images from a camera in case of rain, incident light and contamination | |
| US20240029444A1 (en) | Correction of images from a panoramic-view camera system in the case of rain, incident light and contamination | |
| US20090103779A1 (en) | Multi-sensorial hypothesis based object detector and object pursuer | |
| CN119018043B (en) | Adaptive headlight control method and adaptive headlight system | |
| CN103597498A (en) | Method and control device for transmitting data relating to a current vehicle environment to a headlight control device of a vehicle | |
| CN110371016B (en) | Distance estimation for vehicle headlights | |
| JP2008028957A (en) | Image processing apparatus for vehicle | |
| US12190601B2 (en) | Advanced driver assist system and method of detecting object in the same | |
| CN111323027A (en) | A method and device for making high-precision map based on fusion of lidar and surround-view camera | |
| CN106295493A (en) | Exterior environment recognition device | |
| CN110774975B (en) | Intelligent light control method and device based on image recognition | |
| Ewecker et al. | Provident vehicle detection at night for advanced driver assistance systems | |
| US20230286548A1 (en) | Electronic instrument, movable apparatus, distance calculation method, and storage medium | |
| US12354369B2 (en) | Method for automatically executing a vehicle function, method for evaluating a computer vision method and evaluation circuit for a vehicle | |
| JP2004257837A (en) | Stereo adapter imaging system | |
| CN113632450B (en) | Imaging system and image processing apparatus | |
| Dawid et al. | The system of headlights operation recognition using the digital twin method | |
| Duthon et al. | Quadsight® Vision System in Adverse Weather Maximizing the benefits of visible and thermal cameras | |
| EP3168779A1 (en) | Method for identifying an incoming vehicle and corresponding system | |
| CN117124974A (en) | Anti-dazzle automobile lighting system and automobile |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |