High beam and low beam automatic switching control method and device based on scene recognition

Through the collaborative work of multiple sensors, we can identify target objects around the road environment and vehicles in real time, build scene data sets, and realize automatic switching of high and low beams, which solves the problem of unintelligent switching of high and low beams in the existing technology, and improves the safety of night driving and driving experience.

CN119018045APending Publication Date: 2024-11-26CHONGQING SELIS PHOENIX INTELLIGENT INNOVATION TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411332461.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-24
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing high and low beam switching system only relies on light sensors and cannot effectively deal with complex driving scenarios, resulting in high safety risks of driving at night and cannot intelligently switch high and low beam lights.

Method used

Multi-sensors work together, including light sensors, multi-function cameras, millimeter-wave radars and lidars, use the data fusion module to identify target objects around the road environment and vehicles in real time, build scene data sets, and realize intelligent automatic switching control of high and low beams.

Benefits of technology

It improves the safety and driver experience of driving at night, and can accurately identify and automatically adjust the high and low beam status in a variety of complex driving scenarios to reduce the risk of traffic accidents.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119018045A_ABST
    Figure CN119018045A_ABST
Patent Text Reader

Abstract

The invention provides an automatic switching control method and device for high and low beams based on scene recognition. The method comprises the following steps: identifying a target object by using a first sensor combination to obtain a fusion target confidence coefficient; identifying lane lines in the road environment by using a second sensor combination to obtain a fusion lane line confidence coefficient; determining the size of the target object and the parameters of the lane line, and calculating the speed and relative position relation of the target object according to the size of the target object and the parameters of the lane line; constructing a scene data set according to the illumination intensity, the fusion target confidence, the fusion lane line confidence, the size of the target object, the parameters of the lane line, the speed of the target object and the relative position relationship; and identifying the driving scene of the current vehicle based on the scene data set, and executing a corresponding high and low beam automatic switching control strategy according to the identification result. The high beam and low beam switching can be automatically and effectively carried out according to the road environment, and the intelligence of the system and the driving safety are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of new energy vehicle technology, and in particular to a method and device for automatically switching between high and low beams based on scene recognition. Background Art

[0002] In the existing vehicle high and low beam switching system, a light sensor is usually used to detect the ambient light intensity, and based on this, the vehicle automatically switches to the high beam when the light is dim or at night, and switches to the low beam when the light is sufficient. The system can monitor the ambient light conditions in real time and adjust the high and low beam status of the vehicle.

[0003] However, the current high-beam and low-beam switching system is based on a single detection method of light sensors and cannot fully cope with complex driving scenarios. For example, at night or on roads with poor lighting, when encountering oncoming vehicles, meeting vehicles on a curve, or overtaking large vehicles, if the vehicle keeps the high beam on, it is easy to cause glare to other drivers, thus causing traffic accidents. The limitation of the existing technology is the lack of recognition and response to a variety of driving scenarios, and the inability to achieve intelligent automatic switching of high and low beams in a variety of road environments, which poses a great safety hazard. Summary of the invention

[0004] In view of this, an embodiment of the present application provides a method and device for controlling automatic switching of high and low beams based on scene recognition, so as to solve the problem that the prior art cannot realize intelligent automatic switching of high and low beams according to the road environment, resulting in reduced system intelligence and driving safety.

[0005] In a first aspect of an embodiment of the present application, a method for automatically switching high and low beams based on scene recognition is provided, comprising: obtaining the light intensity of a road environment in which a vehicle is located, identifying a target object in the road environment using a first sensor combination, obtaining multiple target object recognition results and a target confidence corresponding to each target object recognition result; fusing the multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object; identifying lane lines in the road environment using a second sensor combination to obtain multiple lane line recognition results and a lane line confidence corresponding to each lane line recognition result; fusing the multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object; identifying lane lines in the road environment using a second sensor combination to obtain multiple lane line recognition results and a lane line confidence corresponding to each lane line recognition result; fusing the multiple target confidences output by the second sensor combination to obtain a fused target confidence for the target object; The confidences of multiple lane lines obtained are fused to obtain a fused lane line confidence; the size of the target object is determined from the multiple target object recognition results, and the parameters of the lane line are determined from the multiple lane line recognition results, and the speed and relative position relationship of the target object are calculated according to the size of the target object and the parameters of the lane line; a scene data set is constructed according to the light intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed and relative position relationship of the target object; the driving scene in which the current vehicle is located is recognized based on the scene data set, and the corresponding high and low beam automatic switching control strategy is executed according to the recognition result of the driving scene.

[0006] According to a second aspect of an embodiment of the present application, a high and low beam automatic switching control device based on scene recognition is provided, comprising: a first recognition module, configured to obtain the light intensity of the road environment in which the vehicle is located, and use a first sensor combination to identify the target object in the road environment to obtain multiple target object recognition results and a target confidence corresponding to each target object recognition result; a first fusion module, configured to fuse the multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object; a second recognition module, configured to use a second sensor combination to identify the lane line in the road environment to obtain multiple lane line recognition results and a lane line confidence corresponding to each lane line recognition result; and a second fusion module, configured to fuse the multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object. The multiple lane line confidences output by the second sensor combination are fused to obtain a fused lane line confidence; a determination module is configured to determine the size of the target object from the multiple target object recognition results, and determine the parameters of the lane line from the multiple lane line recognition results, and calculate the speed and relative position relationship of the target object according to the size of the target object and the parameters of the lane line; a construction module is configured to construct a scene data set according to the light intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed and relative position relationship of the target object; a control module is configured to identify the driving scene in which the current vehicle is located based on the scene data set, and execute a corresponding high and low beam automatic switching control strategy according to the recognition result of the driving scene.

[0007] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the computer program.

[0008] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0009] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0010] By acquiring the light intensity of the road environment in which the vehicle is located, the target object in the road environment is identified by using the first sensor combination to obtain multiple target object identification results and the target confidence corresponding to each target object identification result; the multiple target confidences output by the first sensor combination are fused to obtain the fused target confidence for the target object; the lane line in the road environment is identified by using the second sensor combination to obtain multiple lane line identification results and the lane line confidence corresponding to each lane line identification result; the multiple lane line confidences output by the second sensor combination are fused to obtain the fused lane line confidence; the size of the target object is determined from the multiple target object identification results, and the parameters of the lane line are determined from the multiple lane line identification results, and the speed and relative position relationship of the target object are calculated according to the size of the target object and the parameters of the lane line; a scene data set is constructed according to the light intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed and relative position relationship of the target object; the driving scene in which the current vehicle is located is identified based on the scene data set, and the corresponding high and low beam automatic switching control strategy is executed according to the recognition result of the driving scene. This application can automatically and effectively switch between high and low beams according to the road environment, improving the intelligence of the system and driving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0012] Figure 1 It is a schematic diagram of the overall implementation process of the high and low beam automatic switching control method based on scene recognition provided in an embodiment of the present application;

[0013] Figure 2 is a schematic diagram of the low beam and high beam switching interval provided in an embodiment of the present application;

[0014] Figure 3 It is a flow chart of a method for automatically switching high and low beams based on scene recognition provided in an embodiment of the present application;

[0015] Figure 4 is a schematic diagram of a multi-sensor based target confidence fusion strategy provided in an embodiment of the present application;

[0016] Figure 5 is a schematic diagram of determining the target vehicle size and distinguishing the vehicle type in an actual scenario provided by an embodiment of the present application;

[0017] Figure 6 This is a schematic diagram of a set of discrete points of lane lines in an actual scenario provided by an embodiment of the present application;

[0018] Figure 7 This is a schematic diagram of automatic switching control of high and low beams of a vehicle in an open road driving scenario provided by an embodiment of the present application;

[0019] Figure 8 It is a schematic diagram of automatic switching control of high and low beams in a scenario where a vehicle is traveling on a straight road or a curve and there is an oncoming vehicle, provided by an embodiment of the present application;

[0020] Fig. 9 It is a schematic diagram of automatic switching control of high and low beams in a scenario where a vehicle is traveling on a straight road or a curve and there is a large vehicle in an adjacent lane, provided by an embodiment of the present application;

[0021] Fig.10 It is a structural schematic diagram of a high and low beam automatic switching control device based on scene recognition provided in an embodiment of the present application;

[0022] Fig.11 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0024] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0025] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc. mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0026] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0027] In the existing high-beam and low-beam switching system, the working principle of the system mainly relies on light sensors. Specifically, when the vehicle is driving in a dark environment (such as at night or in a tunnel), the system will automatically activate the high beam to ensure a clearer view for the driver; and when the vehicle is driving in a well-lit area (such as during the day or on a road with good street lighting), the system will automatically switch to the low beam to avoid glare interference to other road users.

[0028] However, this system has obvious limitations. First, it relies solely on light sensors to determine the intensity of ambient light and to determine when to switch between high and low beams. Although this approach works in some scenarios, it is not suitable for more complex driving environments, such as:

[0029] When driving at night or on a dimly lit road, if there is an oncoming vehicle and the vehicle still keeps the high beam on, it may cause the driver of the oncoming vehicle to be momentarily blinded, increasing the risk of an accident.

[0030] When a vehicle is turning, if there is a vehicle coming from the opposite direction, the high beam may directly shine on the other vehicle, interfering with its driver's vision.

[0031] When a vehicle is overtaking a large vehicle at night or in dim light conditions, if the high beam remains on, it may cause visual interference to the driver of the large vehicle in front, or cause a loss of information about the road conditions in front of the driver before the vehicle completely overtakes.

[0032] Therefore, the existing system cannot effectively switch between high and low beams when encountering these scenarios, which may lead to potential traffic safety hazards. In other words, the disadvantage of the existing technology is that it fails to consider the high and low beam switching requirements in these specific driving scenarios, resulting in the system's intelligence and safety being not perfect.

[0033] In view of the problems existing in the above-mentioned prior art, the present application provides a method for automatically switching high and low beams based on multiple sensor recognition. First, before introducing the technical solution of the present application in detail, the structural composition of the high and low beam automatic switching control system involved in the present application is described. The high and low beam automatic switching control system involved in the present application aims to realize the coordinated work of multiple sensors to perceive and analyze the driving environment around the vehicle in real time, thereby intelligently and accurately controlling the switching of the high and low beams of the vehicle to improve driving safety. The structure of the system and the functions of each component are as follows:

[0034] Light sensor: The light sensor is a basic component of the system, used to detect the light intensity of the vehicle's surrounding environment. The light sensor monitors the light conditions in the environment in real time and measures the current light intensity value (E). When the vehicle is driving in a dark environment (such as at night or in a tunnel), the light sensor will detect a lower light intensity, thereby providing a basis for the automatic high and low beam switching system to decide whether to switch to high beam or low beam.

[0035] Automatic high and low beam switching system: The automatic high and low beam switching system is the core module of the entire control system. It is responsible for intelligently controlling the switching of vehicle lights based on the light intensity data provided by the light sensor and the scene information provided by other sensors. The system not only takes into account the ambient light conditions, but also identifies different driving scenarios (such as oncoming vehicles, meeting on a curve, overtaking large vehicles, etc.) by integrating multiple sensor data, and adjusts the status of the high and low beam lights accordingly to ensure optimal road lighting and safety.

[0036] Multi-function camera: A multi-function camera is a sensor used for image recognition. It can identify and extract lane lines and target objects (such as vehicles and pedestrians) by capturing images of the road. The camera is responsible for identifying the position and shape of the lane lines and providing image data for detecting whether there are oncoming vehicles or other potentially dangerous objects ahead. In addition, the camera can also identify marking lines and traffic signs on the road, providing important reference information for the system's scene judgment.

[0037] Millimeter-wave radar: Millimeter-wave radar detects relevant parameters of surrounding target objects, such as position, speed, and azimuth, by emitting millimeter-wave signals and receiving their reflected signals. Millimeter-wave radar has high accuracy in detecting dynamic information of target objects (such as movement speed and direction), and plays a key role in identifying target objects related to the direction of vehicle travel. It can accurately detect vehicles in the front or adjacent lanes, as well as identify the relative speed and direction of these vehicles, providing the system with real-time environmental perception data.

[0038] LiDAR: LiDAR uses laser beams to scan the surrounding environment, detect and obtain three-dimensional spatial data of target objects, such as distance, direction, height, speed, attitude, shape, etc. LiDAR has advantages in high-precision target detection and tracking, especially in identifying the outline, size and spatial position of the vehicle in front, and can provide accurate data support. The sensor can not only identify stationary objects, but also track moving targets and measure their precise distance and speed, providing basic data for the decision-making of the automatic high and low beam switching system.

[0039] Data fusion module: The data fusion module is the intelligent data processing center in the system, responsible for fusing sensor data from optical sensors, multi-function cameras, millimeter-wave radars and lidars. The module generates a comprehensive scene data set by fusing the recognition results of different sensors, such as target confidence, lane line confidence, target object size, position, speed and lane line curvature. The fusion module performs weighted processing on sensor data to output a more accurate environmental judgment result, providing optimal decision support for the high and low beam switching control system.

[0040] Through the coordinated work of the above-mentioned structure and various components, the high and low beam automatic switching control system of the present application can intelligently identify the current driving environment in a variety of complex driving scenarios, and automatically adjust the vehicle's high and low beam status according to the dynamic changes of ambient lighting and surrounding target objects, thereby greatly improving the safety and driving comfort of night driving.

[0041] Based on the structure of the above-mentioned high and low beam automatic switching control system, the overall implementation process of the high and low beam automatic switching control method based on scene recognition in the present application is summarized below. Figure 1 is a schematic diagram of the overall implementation flow of the high and low beam automatic switching control method based on scene recognition provided in an embodiment of the present application, Figure 2 Schematic diagram of the low beam and high beam switching interval provided in the embodiment of the present application. Figure 1 and Figure 2 As shown, the overall implementation process of the high and low beam automatic switching control method based on scene recognition may specifically include:

[0042] This system uses multi-function cameras, millimeter wave radars, laser radars and other sensors to intelligently identify road types, the presence of oncoming and same-direction vehicles and their relative positions in a variety of driving scenarios, and based on this, it can automatically switch between high and low beams. The following is a detailed description of the overall implementation process of the technical solution of this application.

[0043] like Figure 1 As shown in the figure, the workflow of this system mainly includes the following key steps:

[0044] When the vehicle is in motion and the high and low beam automatic switching control system is turned on, the light sensor will detect the current ambient light intensity E in real time. The function of the light sensor is to monitor the light conditions around the vehicle to determine whether the high beam needs to be turned on or off.

[0045] When the light intensity E is lower than 50lux, the system first determines whether there are other vehicles on the road. If the vehicle is driving on an open road and there are no other vehicles, the system will automatically switch to high beam to improve the driver's field of vision and ensure safety during night driving.

[0046] When the light intensity E is lower than 50lux and the sensor identifies a target vehicle in the opposite lane, the multi-function camera, millimeter-wave radar and lidar will work together to identify the distance, speed and relative position of the opposite vehicle. In order to avoid dazzling the opposite vehicle, the system will automatically switch the high beam to low beam to ensure the driving safety of both parties.

[0047] When the light intensity E is lower than 50lux and the sensor detects a large vehicle (such as a truck or bus) in the adjacent lane, the system will further determine whether the vehicle has the intention to overtake. By monitoring the depth change rate of the driver's accelerator pedal, if the depth change rate of the accelerator pedal is greater than 50%, the system will determine that the driver has the intention to overtake.

[0048] After confirming the intention to overtake, the system will automatically control the headlights to flash twice quickly (such as Figure 2 As shown in the figure, the switching interval between low beam and high beam is 100ms) to remind the drivers of large vehicles in adjacent lanes to pay attention to the overtaking behavior of the vehicle. After flashing the lights, the system will switch from high beam to low beam to ensure safety during overtaking.

[0049] The system of this application not only relies on the single detection of light intensity by the light sensor, but also combines the multi-sensing capabilities of multi-function cameras, millimeter-wave radars and lidars to achieve accurate recognition and processing of complex scenes in the following ways:

[0050] Multi-function camera: Mainly responsible for capturing road images and identifying lane lines, vehicles and other obstacles. Through image processing technology, the camera can detect the specific location of oncoming vehicles or vehicles in front.

[0051] Millimeter-wave radar: By transmitting and receiving millimeter-wave signals, millimeter-wave radar can accurately detect the speed, distance and relative position of vehicles in front and adjacent lanes. This is very important for judging whether oncoming vehicles or large vehicles in adjacent lanes are approaching.

[0052] LiDAR: LiDAR further provides three-dimensional spatial information of the target object, including the shape, height and posture of the vehicle, thereby improving the system's recognition accuracy of the target vehicle.

[0053] Through multi-sensor fusion, the system inputs the acquired data into the data fusion module for processing. The module builds a scene data set based on the current light intensity, road type, the presence of the target vehicle, and the relative speed and position of the vehicle. Based on this data set, the system can identify the current driving scene of the vehicle and make corresponding high and low beam control decisions accordingly:

[0054] On empty night roads with no other vehicles, the system automatically turns on the high beams;

[0055] When detecting oncoming vehicles or large vehicles in adjacent lanes, the system automatically switches to low beam;

[0056] In an overtaking situation, the system alerts the truck driver by flashing lights, and then switches to low beam to ensure the safety of the overtaking process.

[0057] Through the multi-sensor fusion strategy and intelligent scene recognition method of this application, the high and low beam automatic switching control system can respond to various complex driving scenarios more flexibly and accurately. It not only solves the limitations of the single light sensor control method in the prior art, but also realizes the intelligent switching of high and low beam lights through the comprehensive processing of different sensor data, improving the safety of night driving and the driver's experience.

[0058] A vehicle anti-skid control method and device provided according to an embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0059] Figure 3 1 is a flow chart of a method for automatically switching high and low beams based on scene recognition provided by an embodiment of the present application. Figure 3 As shown, the high and low beam automatic switching control method based on scene recognition may specifically include:

[0060] S301, obtaining the light intensity of the road environment where the vehicle is located, identifying the target object in the road environment using the first sensor combination, and obtaining multiple target object recognition results and the target confidence corresponding to each target object recognition result;

[0061] S302, fusing multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object;

[0062] S303, using the second sensor combination to identify lane lines in the road environment, and obtain multiple lane line recognition results and lane line confidences corresponding to each lane line recognition result;

[0063] S304, fusing the multiple lane line confidences output by the second sensor combination to obtain a fused lane line confidence;

[0064] S305, determining the size of the target object from the multiple target object recognition results, determining the parameters of the lane line from the multiple lane line recognition results, and calculating the speed and relative position relationship of the target object according to the size of the target object and the parameters of the lane line;

[0065] S306, constructing a scene data set according to the illumination intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed of the target object and the relative position relationship;

[0066] S307, identifying the driving scene in which the current vehicle is located based on the scene data set, and executing a corresponding high and low beam automatic switching control strategy according to the recognition result of the driving scene.

[0067] In some embodiments, the first sensor combination includes a laser radar, a multi-function camera, and a millimeter wave radar, and multiple target confidences output by the first sensor combination are fused to obtain a fused target confidence for the target object, including:

[0068] Obtaining a first target confidence level for identifying and outputting a target object using a laser radar, obtaining a second target confidence level for identifying and outputting a target object using a multi-function camera, and obtaining a third target confidence level for identifying and outputting a target object using a millimeter-wave radar;

[0069] The first target confidence, the second target confidence and the third target confidence are weighted according to the preset weighting factors corresponding to the different target confidences to obtain the fused target confidence.

[0070] Specifically, this embodiment provides a multi-sensor based target confidence fusion strategy for improving the accuracy of target object recognition in the high and low beam automatic switching control system. This embodiment uses a first sensor combination (including a laser radar, a multi-function camera, and a millimeter wave radar) to comprehensively identify target objects in the road environment, and fuses the recognition results of each sensor to generate a more accurate target confidence, thereby supporting the system's judgment of complex driving scenarios and intelligent switching of high and low beams.

[0071] Figure 4 Schematic diagram of a multi-sensor target confidence fusion strategy provided in an embodiment of the present application. Figure 4As shown in the figure, first, the system uses the laser radar to perform preliminary identification of the target object in the road environment. The laser radar obtains information such as the distance, direction and shape of the target object by emitting a laser beam and receiving its reflected signal. Based on this information, the system calculates a preliminary first target confidence (CL), which indicates the laser radar's recognition reliability of the target object. The first target confidence (CL) is transmitted to the data fusion module as an important input parameter for subsequent fusion calculations.

[0072] Next, the system uses a multifunctional camera to further identify the same target object. The camera captures images of the road, extracts the contour, color and other visual features of the target object, and generates image data related to the target. Based on these image data, the system calculates the second target confidence (CV), which reflects the accuracy of the camera's recognition of the target object. After the camera recognition process is completed, the second target confidence (CV) is also transmitted to the data fusion module.

[0073] Subsequently, the system activates the millimeter-wave radar, which further detects dynamic information such as the position, speed, and azimuth of the target object by emitting millimeter-wave signals and receiving the reflected signals. The millimeter-wave radar performs particularly well in bad weather and low-light conditions and can provide additional confidence support. By processing the signal returned by the millimeter-wave radar, the system calculates the third target confidence (CR), which represents the reliability of the millimeter-wave radar's recognition of the target object, and transmits it to the data fusion module.

[0074] Furthermore, in this embodiment, the data fusion module is responsible for fusing the confidences from the laser radar, the multi-function camera and the millimeter wave radar. Specifically, the data fusion module first receives the first target confidence (CL), the second target confidence (CV) and the third target confidence (CR), and then performs weighted calculation on each confidence according to the preset weighting factors of different sensors.

[0075] In practical applications, the weighting factor is set based on the reliability of the sensor in different scenarios. For example, when the ambient lighting conditions are poor, the weighting factor of the millimeter-wave radar may be relatively high because it can still provide stable recognition information under low light conditions. In good lighting conditions, the weighting factor of the camera may be high to make full use of its high sensitivity to visual features.

[0076] The following is a specific example to illustrate the implementation process of target confidence fusion. In one example, for vehicle and other object recognition, when the laser radar recognizes the target, the output target confidence is CL, the camera recognizes the target output confidence is CV, and the millimeter wave radar recognizes the target output confidence is CR. The fusion module performs weighted calculation based on the confidence of each target:

[0077]

[0078] Among them, CT1 represents the fusion target confidence output by the fusion module. When CT1 is greater than 0.7, the target confidence is considered to be high, and the output is the determined actual target. KL represents the lidar weighting factor, KV represents the multi-function camera weighting factor, KR represents the millimeter-wave radar weighting factor, CL represents the first target confidence, CV represents the second target confidence, and CR represents the third target confidence.

[0079] Furthermore, after the fusion process is completed, the data fusion module outputs the calculated fusion target confidence (CT) to the system for judgment and analysis of the current scene. In this embodiment, the system detects and identifies environmental targets once per second to ensure that the surrounding target objects and scene changes can still be accurately identified when the vehicle is traveling at high speed.

[0080] Furthermore, based on the fused target confidence (CT), the system accurately classifies and judges the target objects in the current scene, ensuring that in complex scenarios such as oncoming vehicles, meeting vehicles on curves, and overtaking by the vehicle itself, the high and low beams can be switched in time to avoid dazzling other road users, thereby ensuring driving safety.

[0081] Through the method of the above-mentioned embodiment, the multi-sensor fusion strategy of the present application effectively improves the accuracy and reliability of target recognition, especially in complex driving environments, and realizes intelligent switching of high and low beam headlights through precise scene judgment, further improving the safety of night driving and driving experience.

[0082] In some embodiments, the second sensor combination includes a laser radar and a multi-function camera, and multiple lane line confidences output by the second sensor combination are fused to obtain a fused lane line confidence, including:

[0083] Obtaining a first lane line confidence value output by using a laser radar to identify a lane line in a road environment;

[0084] Obtaining a second lane line confidence value output by recognizing a lane line in a road environment using a multifunctional camera;

[0085] The first lane line confidence and the second lane line confidence are weightedly calculated according to the preset weighting factors corresponding to the different lane line confidences to obtain the fused lane line confidence.

[0086] Specifically, lane line recognition is crucial for vehicles to maintain the correct lane and path planning. Different sensors have their own advantages and limitations when identifying lane lines. Among them, LiDAR can measure the distance, shape and position of lane lines by emitting laser pulses and receiving their reflected signals. LiDAR is good at working under various lighting conditions and is not interfered by external light, especially at night or in low visibility conditions. Multi-function cameras mainly capture road images and then extract lane line information through image processing technology. Cameras are very effective in detecting visual features such as color and width of lane lines, but their performance may decline in poor lighting conditions (such as at night, rainy and foggy weather).

[0087] In order to improve the accuracy and reliability of lane line recognition, the embodiment of the present application adopts a lane line confidence fusion strategy based on a second sensor combination (including a laser radar and a multi-function camera). The specific implementation process of this embodiment is as follows:

[0088] First, the LiDAR detects the road surface to identify the position and shape of the lane line and outputs a confidence level (CL). This confidence level represents the confidence level of the LiDAR in the recognition result. The higher the value, the more reliable the LiDAR is in recognizing the lane line.

[0089] Next, after the LiDAR recognition, the multi-function camera will recognize the same lane line. The multi-function camera uses image processing algorithms to extract the visual features of the lane line (such as color, shape, width, etc.) and outputs a confidence level (CV) that indicates the multi-function camera's confidence in the recognition result.

[0090] Finally, after obtaining the lane line confidence output by the lidar and the multi-function camera respectively, the system's fusion processing module will perform weighted fusion on the two lane line confidences to obtain a comprehensive lane line confidence (CT2).

[0091] For example, in one example, for lane line recognition calculation, since different sensors have different characteristics for target recognition, this embodiment only performs lane line confidence fusion calculation on the laser radar and the multi-function camera:

[0092]

[0093] Among them, CT2 represents the fused lane line confidence output by the fusion module, KL represents the lidar weighting factor, KV represents the multi-function camera weighting factor, CL represents the first lane line confidence, and CV represents the second lane line confidence.

[0094] It should be noted that when the integrated lane line confidence (i.e., fused lane line confidence) CT2 is greater than 0.6, the system considers the lane line recognition result to be reliable and outputs the determined actual lane line target. In other words, the vehicle control system can rely on this recognition result to determine the position of the current lane line, and thus perform corresponding driving operations, such as keeping the lane, adjusting the direction, etc.

[0095] Through the method of the present embodiment described above, this lane line confidence fusion strategy combines the respective advantages of the laser radar and the multi-function camera, thereby improving the accuracy and reliability of lane line recognition. By combining the recognition results of the two sensors, the system can more accurately identify lane lines, especially in complex environments, such as lighting changes, road marking wear, etc., the fusion strategy can provide more stable recognition results. Relying on a single sensor alone may cause recognition errors due to environmental changes, and by fusing the confidence of multiple sensors, such misidentification can be reduced and the overall reliability of the system can be improved.

[0096] In some embodiments, before determining the size of the target object from the plurality of target object recognition results, the method further comprises:

[0097] The fusion target confidence is compared with the preset fusion target confidence threshold. When the fusion target confidence exceeds the fusion target confidence threshold, the target object is determined as the actual target; the fusion lane line confidence is compared with the preset fusion lane line confidence threshold. When the fusion lane line confidence exceeds the fusion lane line confidence threshold, the lane line is determined as the actual lane line.

[0098] Specifically, this embodiment also proposes a target object and lane line recognition method based on a fusion confidence threshold. Before determining the size of the target object and the lane line parameters, the method first compares the fusion confidence with a preset threshold to determine the reliability of the recognition result. Only when the confidence exceeds the preset threshold will the system determine the target object or lane line recognition result as an actual valid recognition result, thereby ensuring the reliability and safety of the vehicle control system.

[0099] In one or more embodiments, the system first calculates the fused target confidence (CT1) of the target object through a multi-sensor fusion method. This confidence integrates the recognition results of the lidar, multi-function camera and millimeter-wave radar, and reflects the overall reliability of the system's recognition of the target object.

[0100] To ensure the accuracy of the recognition result, the system compares CT1 with the preset fusion target confidence threshold. In one example, the fusion target confidence threshold can be set to 0.7. When CT1 is greater than 0.7, the system considers that the recognition result of the target object has a high degree of confidence, and at this time, the target object is determined to be an actual target, that is, a effectively identified vehicle or obstacle.

[0101] For example, if the CT1 value obtained after the weighted calculation of the LiDAR, multi-function camera and millimeter-wave radar is 0.75, since this value is higher than the preset threshold of 0.7, the system will confirm the target object recognition result as an actual target vehicle or obstacle. This confirmation step ensures that in the case of possible recognition errors, only recognition results with high enough confidence will be adopted to prevent wrong judgments from affecting driving safety.

[0102] After the target object recognition is verified, the system performs similar verification processing on the lane line recognition results. Through the multi-sensor fusion method, the system calculates the fused lane line confidence (CT2), which represents the overall reliability of the system's lane line recognition.

[0103] In one or more embodiments, the system compares CT2 with a preset fused lane line confidence threshold. In one example, the fused lane line confidence threshold can be set to 0.6. When CT2 is greater than 0.6, the system considers the lane line recognition result reliable and outputs it as the actual lane line position.

[0104] For example, after the system performs lane line recognition, the fused lane line confidence CT2 obtained is 0.65. Since this confidence is higher than the threshold of 0.6, the system will confirm that the lane line recognition result is valid and use it for subsequent driving operations, such as lane keeping and steering adjustment.

[0105] Furthermore, after the confidence verification is passed, the system performs corresponding vehicle control operations according to the effective recognition results. For example, once the confidence of the target object exceeds the threshold and is confirmed as an actual target, the system will perform operations such as adjusting the vehicle speed and avoiding obstacles based on the size, position and speed information of the target object. For example, when a high-confidence vehicle is detected in front, the system may slow down and adjust the high and low beam lights to avoid dazzling oncoming vehicles.

[0106] For another example, after the lane line confidence verification is passed, the system will perform operations such as keeping the vehicle in the center of the lane or adjusting the direction based on the position and curvature information of the lane line. For example, on a highway, when the system recognizes that the lane line has a high confidence level and is a straight line, the system will ensure that the vehicle drives steadily in a straight line and adjust the direction if necessary to avoid deviating from the lane.

[0107] Through the above-mentioned method of this embodiment, by verifying the confidence of the recognition results before actual operation, this embodiment greatly improves the reliability and accuracy of the system, ensures that the vehicle can make safe and reasonable decisions in complex driving environments, and further improves the driver's driving experience and driving safety.

[0108] In some embodiments, this embodiment also provides a method for determining the size of a target object, which is particularly suitable for classifying and identifying target vehicles in a high and low beam automatic switching control system. The method uses a camera and a laser radar to accurately measure the external dimensions of the target vehicle, and classifies the target vehicle into different types based on the measurement results, thereby providing support for subsequent driving decisions.

[0109] When the vehicle is driving, the system first uses the multi-function camera and laser radar to identify the target object on the road ahead. When the target object is identified as a vehicle, the system further measures the size of the vehicle. Figure 5 Schematic diagram of determining the target vehicle size and distinguishing the vehicle type in an actual scenario provided by the embodiment of the present application. Figure 5 As shown, the specific implementation process is as follows:

[0110] The multifunctional camera captures the image of the target vehicle and uses the image processing algorithm to extract the outer contour information of the vehicle. The camera is able to recognize preliminary information about the length, width and height of the vehicle.

[0111] LiDAR emits laser beams and receives reflected signals to accurately measure the distance between the target vehicle and the vehicle, as well as the three-dimensional shape of the target vehicle. This information is used to supplement and verify the image data obtained by the camera.

[0112] After the camera and lidar complete preliminary recognition and data collection, the system preprocesses the target object to form a three-dimensional target box that contains the main size parameters of the target vehicle, such as length (D), width (L) and height (H).

[0113] Based on the target frame obtained after preprocessing, the system automatically calculates the three key dimensions of the target vehicle: Length (D) represents the dimension of the target vehicle in the direction of travel, that is, the front-to-back length of the vehicle. Width (L) represents the lateral dimension of the target vehicle, that is, the left-right width of the vehicle. Height (H) represents the vertical height of the target vehicle from the ground to the roof.

[0114] The system uses the distance and height data provided by the LiDAR, combined with the image information from the camera, to accurately calculate the above three size parameters through the sensor algorithm. Through this multi-sensor fusion method, the system can more accurately determine the specific size of the target vehicle.

[0115] After determining the size parameters of the target vehicle, this embodiment classifies the target vehicle by size according to its length (D):

[0116] If the length D of the target vehicle is greater than 6 meters, the system classifies the target vehicle as a large vehicle. Large vehicles usually include trucks, buses and other larger vehicles. Such vehicles are marked as tp=1 in the dataset.

[0117] If the length D of the target vehicle is less than or equal to 6 meters, the system classifies the target vehicle as a small car. Small cars usually include sedans, SUVs, etc. This type of vehicle is marked as tp=2 in the dataset.

[0118] This classification result is particularly critical in the high and low beam switching control of the subsequent embodiments. For example, in an overtaking scenario, if the identified target vehicle is a large vehicle (tp=1), the system may adopt different light switching strategies to ensure the safety of the overtaking process and the visual comfort of other drivers.

[0119] In practical applications, the target object size determination method of this embodiment can greatly improve the system's recognition accuracy of the vehicle ahead. For example, when the system recognizes that there is a truck ahead, whose length D is 8 meters (greater than 6 meters), the system will mark the target vehicle as a large vehicle (tp=1). At this time, the system will adopt corresponding lighting control strategies based on this mark, which may include automatically flashing lights to remind when overtaking, and avoiding unnecessary glare to the driver of the large vehicle.

[0120] For another example, when the vehicle ahead is identified as a sedan (small car, tp=2) with a length of 4.5 meters, the system may adopt different high and low beam switching strategies to adapt to the sight height and reaction speed of the small car driver.

[0121] Through the method of the above-mentioned embodiment, through this multi-sensor fusion target size recognition and classification method, this embodiment ensures that the high and low beam automatic switching control system can adapt to the driving scenarios of different types of vehicles, thereby improving the vehicle's driving safety and the driver's experience.

[0122] In some embodiments, determining lane line parameters from a plurality of lane line recognition results includes:

[0123] Process the road image to extract pixel data containing lane line information, and collect point cloud data on the road surface to obtain discrete point data containing lane line information;

[0124] The pixel point data and the discrete point data are processed separately to obtain a discrete point set of the lane line, and the discrete point set is merged and fitted to generate the width and curvature parameters of the lane line;

[0125] The curvature direction of the lane line is determined according to the curvature parameter, and it is judged whether the lane line is a straight lane.

[0126] Specifically, this embodiment provides a lane line parameter determination method based on multi-sensor data fusion, which is particularly suitable for accurate recognition of lane line shape in a high and low beam automatic switching control system. The method collects road images and point cloud data through a camera and a laser radar respectively, and uses a fusion algorithm to calculate the width and curvature parameters of the lane line, thereby helping the system to determine the curvature direction of the lane line and whether the lane line is a straight lane.

[0127] In one or more embodiments, the system first processes the road image through a multifunctional camera to extract pixel data containing lane line information. These pixel data include information such as the position, shape, and color of the lane line in the image, and are an important data source for preliminary lane line recognition.

[0128] At the same time, the system also collects point cloud data of the road surface through LiDAR. LiDAR emits laser pulses and detects reflected signals to obtain three-dimensional point cloud data of the road surface. These point cloud data can accurately reflect the height, depth and distribution of lane lines in three-dimensional space.

[0129] Furthermore, after the camera and lidar collect pixel data and point cloud data respectively, the system will preliminarily process these data to extract discrete point information related to the lane line. Among them, pixel data processing includes the system performing pixel processing on the image captured by the camera, extracting the lane line related pixels from the image, and forming a set of two-dimensional discrete point data sets. Point cloud data processing includes the system analyzing the point cloud data collected by the lidar, extracting the three-dimensional discrete points corresponding to the lane line, and forming a three-dimensional discrete point data set.

[0130] Next, the system merges these two-dimensional and three-dimensional discrete point data sets to obtain a unified lane line discrete point set. Figure 6 Schematic diagram of a set of lane line discrete points in an actual scenario provided by an embodiment of the present application. Figure 6 As shown, the discrete point set combines information from different sensors to ensure that the lane line recognition is more accurate and comprehensive.

[0131] Furthermore, after obtaining the discrete point set of the lane line, the system performs curve fitting on these discrete points to generate the width and curvature parameters of the lane line. The specific implementation process is as follows:

[0132] The system uses mathematical fitting methods such as the least squares method to fit a set of discrete points and generate a curve equation for the lane line. This process takes into account the spatial distribution of the lane line and fits the set of discrete points into a continuous curve. Through this fitting method, the system can accurately calculate the shape of the lane line and its actual position on the road.

[0133] Based on the lane line curve equation obtained by fitting, the system further calculates the lane line width and curvature parameters. The width parameter indicates the distance between lane lines, reflecting the width of the road. The curvature parameter indicates the curvature of the lane line. The greater the curvature, the more obvious the lane line curvature.

[0134] By analyzing the curvature parameter, the system can determine the curvature direction of the lane line and whether it is a straight lane: if the curvature parameter is a positive value, the system determines that the lane line is curved to the left; if the curvature parameter is a negative value, the lane line is curved to the right. When the curvature parameter is close to zero, the system determines that the lane line is close to a straight line. In actual applications, the system compares the curvature parameter with the set threshold. If the absolute value of the curvature parameter is less than a very small value (such as 10^-5), the system considers the lane line to be a straight lane. This judgment is very important to ensure that the vehicle maintains stable driving on a straight road.

[0135] In the high and low beam automatic switching control system, the lane line parameter determination method of this embodiment can greatly improve the system's adaptability to complex roads. For example, when a vehicle is driving on a curve, the system calculates the curvature of the lane line to determine the direction and degree of the lane's curvature, thereby adjusting the state of the high and low beam lights to avoid dazzling oncoming vehicles.

[0136] Similarly, when the system recognizes that the vehicle is in a straight lane, it can ensure that the vehicle maintains straight driving with appropriate lighting settings and driving conditions, reducing the possibility of accidents.

[0137] Through the method of the above-mentioned embodiment, the lane line parameter determination method of the present application not only improves the accuracy of lane line recognition, but also provides reliable data support for subsequent driving control operations, ensuring that the intelligent switching of high and low beam headlights can operate stably under various complex road conditions.

[0138] In some embodiments, the discrete point set is merged and fitted to generate lane line width and curvature parameters, including:

[0139] According to the coordinate data set of the discrete points of the lane line, the least square method is used to perform curve fitting on the discrete points to obtain the lane line equation; based on the lane line equation, the first-order derivative and the second-order derivative of the lane line are calculated, and the curvature and curvature radius of the lane line are calculated based on the first-order derivative and the second-order derivative;

[0140] Among them, the curvature is used to characterize the degree of curvature of the lane line. When the curvature is a positive value, it means that the lane is curved to the left; when the curvature is a negative value, it means that the lane is curved to the right; when the curvature is close to 0, it means that the lane line is a straight line.

[0141] Specifically, this embodiment also provides a method for merging and fitting a set of discrete points of lane lines based on the least squares method. The method is used to generate lane line width and curvature parameters to help the system determine the curvature and direction of the lane lines, thereby optimizing the vehicle's high and low beam control strategy.

[0142] This embodiment first collects the discrete point data of the lane line through the camera and the laser radar. The camera is responsible for capturing the image information of the road, and the image processing algorithm is used to extract the two-dimensional pixel coordinates (xi, yi) related to the lane line. These pixel coordinates represent the specific position of the lane line in the image.

[0143] Furthermore, the LiDAR scans the road surface to obtain three-dimensional point cloud data reflecting the height and depth of the lane line. The system aligns and integrates these point cloud data with the two-dimensional data obtained by the camera to form a discrete point set containing lane line information.

[0144] Furthermore, after obtaining a set of discrete points of the lane line, this embodiment performs curve fitting on these discrete points by using the least square method to generate a mathematical equation of the lane line.

[0145] Assume that the discrete point set of the lane line is represented as [(x1,y1),(x2,y2),…,(xn,yn)], and these coordinate points reflect the actual distribution of the lane line on the road surface. In order to find a curve that best fits these discrete points, the system applies the least squares method. The specific fitting goal is to minimize the sum of squared errors between the fitting curve and the actual discrete points through the following formula:

[0146]

[0147] Through the above formula, the system can fit the lane line equation y=f(x), which represents the overall shape of the lane line on the road.

[0148] After obtaining the lane line equation y=f(x), the system further calculates the curvature K and curvature radius R of the lane line to quantify the curvature degree and direction of the lane line.

[0149] The system first calculates the first-order derivative y′=f′(x) and the second-order derivative y″=f″(x) of the lane line equation. The first-order derivative represents the slope of the lane line at a certain point, while the second-order derivative represents the rate of change of the curvature of the lane line at that point. Based on the first-order derivative and the second-order derivative, the system calculates the curvature K of the lane line using the following formula:

[0150]

[0151] The curvature K is used to characterize the curvature of the lane line. If K>0, it means that the lane is curved to the left; if K<0, it means that the lane is curved to the right; if K=0, it means that the lane line is a straight line.

[0152] The system further calculates the radius of curvature R, and the formula is:

[0153]

[0154] The radius of curvature R reflects the bending radius of the lane line. The larger the radius, the straighter the lane line, and the smaller the radius, the more curved the lane line.

[0155] After calculating the curvature K and the curvature radius R of the lane line, this embodiment further determines the curvature direction and straight line characteristics of the lane line:

[0156] By analyzing the sign of the curvature K, the system can determine the direction of the lane line. When K>0, the lane line bends to the left; when K<0, the lane line bends to the right. This information is crucial for the vehicle's driving stability in a curve.

[0157] If the curvature K is close to zero (for example, when K is on the order of 10 -5 ), the system will consider the lane line to be a straight line. At this point, the vehicle can continue to maintain its current direction without making significant steering adjustments.

[0158] In the high and low beam automatic switching control system, the lane line parameter determination method of this embodiment can provide the system with accurate lane line information, especially in complex road environments, such as mountain roads with many bends or straight sections of highways. By accurately calculating the width and curvature of the lane line, the system can dynamically adjust the vehicle's lighting status in different driving scenarios to ensure the driver's clear field of vision while avoiding dazzling interference to other road users.

[0159] For example, when the system recognizes that the vehicle is about to enter a left turn (curvature K>0), the system will adjust the high and low beams in advance to adapt to the lighting requirements of the curve; on a straight section (curvature K≈0), the system can continue to keep the high beams on to provide better lighting.

[0160] Through this precise lane curvature recognition and calculation method, this embodiment significantly improves the driving safety and comfort of the vehicle under various road conditions.

[0161] In some embodiments, the present embodiment will also construct a scene data set based on the illumination intensity, fused target confidence, fused lane line confidence, target object size, lane line parameters, target object speed and relative position relationship. The following is a detailed description of the construction process of the scene data set with specific examples, which may include the following contents:

[0162] Through the ambient light intensity E, the target vehicle speed V and direction, - indicates that the target vehicle is traveling in the opposite direction, + indicates that the target vehicle is traveling in the same direction, the relative position relationship Q, - indicates that the target vehicle is on the left side of the vehicle, + indicates that the target vehicle is on the right side of the vehicle, when the target vehicle and the vehicle are in the same lane, Q = 0, the longitudinal relative distance between the vehicle and the target vehicle is Z, and the target confidence calculated by the fusion module, the lane line curvature can form a scene data set M:

[0163]

[0164] When the target vehicle speed V is a negative value, the target relative position Q can only take a negative value; the current scene of the vehicle is confirmed by judging the parameters of the data set.

[0165] For example, in an actual driving scenario, the scenario dataset generated by the system is: This dataset is reflected in the scene as the current ambient light intensity of 30lux, the sensor fusion output target confidence of 0.8, and the lane line confidence of 0.7, which are high confidence levels. The target is considered to be real. On a right curve with a curvature of 0.002 (curve radius of 500m), in the right lane, there is a large truck traveling in the same direction at a lateral distance of 2m from the vehicle. The target size is 8×2.4×3m and the longitudinal distance from the vehicle is 70m. At the same time, combined with the change rate of the accelerator pedal depth of the vehicle, when the vehicle intends to overtake the large truck, the automatic high and low beam switching system will control the vehicle's headlights to flash twice and then switch to low beam.

[0166] In some embodiments, according to the recognition result of the driving scene, the corresponding high and low beam automatic switching control strategy is executed, including:

[0167] When the driving scene is that the current vehicle is driving on a straight road or a curve, the light intensity meets the preset high beam on condition, and when a target vehicle is identified in the opposite lane, the current vehicle is automatically controlled to turn off the high beam and switch to the low beam;

[0168] When the driving scenario is that the current vehicle is driving on a straight road or a curve, the light intensity meets the preset high beam on condition, and when a target vehicle is identified in the adjacent lane and it is determined that the current vehicle has the intention to overtake, the headlights of the current vehicle are automatically controlled to flash and switch to low beam.

[0169] Specifically, this embodiment also proposes a high and low beam automatic switching control strategy based on the driving scene recognition result, which can dynamically adjust the high and low beam status of the vehicle according to different driving scenarios to ensure driving safety. In particular, when the system recognizes an oncoming vehicle or overtaking scene, the system can automatically perform the corresponding light switching operation to avoid dazzling other road users and ensure the safety of vehicle driving.

[0170] The system first detects the light intensity E around the vehicle in real time through the light sensor. When the light intensity is lower than the preset high beam on condition (for example, light intensity E < 50 lux), the system will consider that the current ambient light is insufficient and the high beam needs to be turned on to provide better lighting.

[0171] At the same time, the system uses multi-function cameras, millimeter-wave radars and lidars to identify the current driving scene of the vehicle in real time. The system can identify whether the vehicle is on a straight road or a curve, and detect whether there are other vehicles around (such as vehicles coming from the opposite lane or large vehicles in the adjacent lane).

[0172] The following is a detailed description of the high and low beam automatic switching control strategy based on the driving scene recognition result of this embodiment in conjunction with specific scenarios and drawings, which may specifically include the following contents:

[0173] Driving scene 1

[0174] Figure 7 Schematic diagram of automatic switching control of high and low beams of a vehicle in an open road driving scenario provided by an embodiment of the present application. Figure 7 As shown, when the light intensity E meets the high beam turning-on condition (for example, when the light intensity E is lower than 50 lux), and the current vehicle is traveling on an open road with no other vehicles, it automatically switches to the high beam.

[0175] Driving scene 2

[0176] Figure 8 1 is a schematic diagram of the automatic switching control of high and low beams in the scenario of a vehicle driving on a straight road or a curve and an oncoming vehicle provided by an embodiment of the present application. Figure 8 As shown in the figure, when the system detects that the vehicle is driving on a straight road or a curve, and the light intensity E meets the high beam on condition, the system will further determine whether there is a target vehicle coming in the opposite lane. The specific implementation process is as follows:

[0177] Through sensor detection, if the system identifies that a target vehicle is approaching in the opposite lane (for example, the camera captures the headlights of the opposite vehicle or the radar detects that the distance to the opposite vehicle is shortened), the system will immediately switch to the high beam.

[0178] To avoid dazzling oncoming drivers, the system automatically turns off the high beam and switches to low beam. This ensures that in low-light conditions, the road is still adequately illuminated while minimizing visual disturbance to oncoming drivers.

[0179] For example, when the light intensity E is detected as 40 lux, the system recognizes through the camera that a car is approaching in the opposite lane, and the system will automatically switch the vehicle's high beam to low beam to ensure the driving safety of both vehicles.

[0180] Driving scene three

[0181] Fig. 9 Schematic diagram of automatic switching control of high and low beams in a scenario where a vehicle is driving on a straight road or a curve and there is a large vehicle in an adjacent lane, provided by an embodiment of the present application. Fig. 9 As shown in the figure, the vehicle's high and low beam control strategy in the overtaking scenario. When the system recognizes that the vehicle has the intention to overtake and the light intensity meets the high beam on condition, the system will perform the following operations:

[0182] First, overtaking intention recognition: By monitoring the driver's operating behavior (for example, the rate of change in the depth of the accelerator pedal is greater than 50%), the system can determine whether the current vehicle has the intention to overtake. At the same time, the system uses millimeter-wave radar and lidar to identify whether there are large vehicles (such as trucks or buses) in the adjacent lane.

[0183] Secondly, flashing lights to warn: When the system determines that the vehicle intends to overtake and there is a large vehicle in the adjacent lane, the system will automatically control the vehicle headlights to flash twice (i.e. quickly switch between high and low beams, with a switching interval of 100ms) to remind the driver of the large vehicle in the adjacent lane that the current vehicle is about to overtake. This operation helps to increase the alertness of the driver of the large vehicle and reduce potential risks during overtaking.

[0184] Finally, automatically switch to low beam: After completing the headlight flashing, the system will automatically switch the high beam to low beam to ensure that there is no continuous dazzle to the driver of the large vehicle during overtaking, while providing sufficient road lighting.

[0185] For example, when the light intensity E is 30lux, the system detects that the depth change rate of the vehicle's accelerator pedal reaches 60%, and recognizes that there is a truck driving in the adjacent lane. The system will automatically flash the headlights and switch to low beam when overtaking begins to ensure the safety of the overtaking process.

[0186] Through the above control strategy, this embodiment can dynamically adjust the status of the vehicle's high and low beams according to the real-time recognized driving scene. At night or in an environment with insufficient light, the system can intelligently switch between high and low beams to adapt to different driving scenarios, ensuring the driver's driving safety and the visual comfort of other road users. The high and low beam automatic switching control strategy of this embodiment enables the system to respond to complex nighttime driving scenarios more intelligently, further improving the vehicle's automation control level and driving safety.

[0187] According to the technical solution provided in the embodiments of the present application, the technical solution of the present application has at least the following advantages:

[0188] This application effectively improves the ability to identify complex driving scenarios by comprehensively utilizing multiple sensors such as multi-function cameras, millimeter-wave radars, and lidars. Traditional high and low beam automatic switching systems often rely on only a single light sensor, but this application can accurately identify more key scenarios, such as oncoming vehicles, meeting vehicles on curves, and overtaking in the same direction, through multi-sensor fusion. In these scenarios, the system can automatically switch between high and low beams based on real-time environmental information, thereby avoiding dazzling interference caused by high beam illumination to drivers of oncoming or adjacent vehicles, effectively reducing the risk of traffic accidents, and significantly improving the driving safety of drivers.

[0189] This application proposes a multi-sensor target fusion control method, which significantly improves the accuracy and stability of target recognition by fusing the confidence of the target recognized by each sensor. Through the preset confidence threshold, this application can effectively filter out false recognition, ensuring that the system only performs light switching operations when the target is clear and the credibility is high, avoiding false operations caused by wrong recognition. Such a design ensures the accuracy of the system when making scene judgments, making the switching of high and low beams more intelligent and reliable, and will not erroneously trigger light switching due to the absence of target objects or the presence of false targets in the environment, thereby further improving the safety and comfort of the overall driving experience.

[0190] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0191] Fig.10 Schematic diagram of the structure of the high and low beam automatic switching control device based on scene recognition provided by the embodiment of the present application. Fig.10 As shown, the high and low beam automatic switching control device based on scene recognition includes:

[0192] The first recognition module 1001 is configured to obtain the light intensity of the road environment where the vehicle is located, identify the target object in the road environment using the first sensor combination, and obtain multiple target object recognition results and the target confidence corresponding to each target object recognition result;

[0193] A first fusion module 1002 is configured to fuse multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object;

[0194] The second recognition module 1003 is configured to recognize lane lines in the road environment using a second sensor combination to obtain a plurality of lane line recognition results and a lane line confidence corresponding to each lane line recognition result;

[0195] A second fusion module 1004 is configured to fuse the multiple lane line confidences output by the second sensor combination to obtain a fused lane line confidence;

[0196] The determination module 1005 is configured to determine the size of the target object from the multiple target object recognition results, determine the parameters of the lane line from the multiple lane line recognition results, and calculate the speed and relative position relationship of the target object according to the size of the target object and the parameters of the lane line;

[0197] A construction module 1006 is configured to construct a scene data set according to the illumination intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed of the target object and the relative position relationship;

[0198] The control module 1007 is configured to identify the driving scene in which the current vehicle is located based on the scene data set, and execute a corresponding high and low beam automatic switching control strategy according to the recognition result of the driving scene.

[0199] In some embodiments, the first sensor combination includes a laser radar, a multi-function camera, and a millimeter-wave radar. Fig.10 The first fusion module 1002 obtains a first target confidence for identifying and outputting a target object using a laser radar, obtains a second target confidence for identifying and outputting a target object using a multi-function camera, and obtains a third target confidence for identifying and outputting a target object using a millimeter-wave radar; and performs weighted calculation on the first target confidence, the second target confidence, and the third target confidence according to weighting factors corresponding to different preset target confidences to obtain the fused target confidence.

[0200] In some embodiments, the second sensor combination includes a laser radar and a multi-function camera, Fig.10The second fusion module 1004 obtains the first lane line confidence which is output by identifying the lane line in the road environment using the laser radar; obtains the second lane line confidence which is output by identifying the lane line in the road environment using the multi-function camera; performs weighted calculation on the first lane line confidence and the second lane line confidence according to the weighting factors corresponding to the preset different lane line confidences to obtain the fused lane line confidence.

[0201] In some embodiments, Fig.10 The determination module 1005 compares the fused target confidence with a preset fused target confidence threshold before determining the size of the target object from the multiple target object recognition results, and when the fused target confidence exceeds the fused target confidence threshold, determines the target object as an actual target; and compares the fused lane line confidence with a preset fused lane line confidence threshold, and when the fused lane line confidence exceeds the fused lane line confidence threshold, determines the lane line as an actual lane line.

[0202] In some embodiments, Fig.10 The determination module 1005 processes the road image to extract pixel point data containing lane line information, and collects point cloud data of the road surface to obtain discrete point data containing lane line information; the pixel point data and the discrete point data are processed separately to obtain a discrete point set of the lane line, and the discrete point set is merged and fitted to generate the width and curvature parameters of the lane line; the curvature direction of the lane line is determined according to the curvature parameters, and it is judged whether the lane line is a straight lane.

[0203] In some embodiments, Fig.10 The determination module 1005 performs curve fitting on the discrete points of the lane line using the least square method according to the coordinate data set of the discrete points of the lane line to obtain the lane line equation; based on the lane line equation, the first-order derivative and the second-order derivative of the lane line are calculated, and the curvature and the curvature radius of the lane line are calculated based on the first-order derivative and the second-order derivative; wherein the curvature is used to characterize the curvature of the lane line, when the curvature is a positive value, it indicates that the lane is curved to the left; when the curvature is a negative value, it indicates that the lane is curved to the right; when the curvature is close to 0, it indicates that the lane line is a straight line.

[0204] In some embodiments, Fig.10The control module 1007 is used to automatically control the current vehicle to turn off the high beam and switch to the low beam when the driving scenario is that the current vehicle is driving on a straight road or a curve, the light intensity meets the preset high beam on condition, and a target vehicle is identified in the opposite lane; when the driving scenario is that the current vehicle is driving on a straight road or a curve, the light intensity meets the preset high beam on condition, a target vehicle is identified in the adjacent lane, and it is determined that the current vehicle has the intention to overtake, the headlights of the current vehicle are automatically controlled to flash and switch to the low beam.

[0205] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0206] Fig.11 1 is a schematic diagram of the structure of the electronic device 11 provided in the embodiment of the present application. Fig.11 As shown, the electronic device 11 of this embodiment includes: a processor 1101, a memory 1102, and a computer program 1103 stored in the memory 1102 and executable on the processor 1101. When the processor 1101 executes the computer program 1103, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 1101 executes the computer program 1103, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0207] Exemplarily, the computer program 1103 may be divided into one or more modules / units, which are stored in the memory 1102 and executed by the processor 1101 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 1103 in the electronic device 11.

[0208] The electronic device 11 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 11 may include, but is not limited to, a processor 1101 and a memory 1102. Those skilled in the art will appreciate that Fig.11 It is only an example of the electronic device 11 and does not constitute a limitation of the electronic device 11. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device may also include input and output devices, network access devices, buses, etc.

[0209] The processor 1101 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0210] The memory 1102 may be an internal storage unit of the electronic device 11, for example, a hard disk or memory of the electronic device 11. The memory 1102 may also be an external storage device of the electronic device 11, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card (FlashCard), etc. equipped on the electronic device 11. Further, the memory 1102 may also include both an internal storage unit of the electronic device 11 and an external storage device. The memory 1102 is used to store computer programs and other programs and data required by the electronic device. The memory 1102 may also be used to temporarily store data that has been output or is to be output.

[0211] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0212] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0213] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0214] In the embodiments provided in the present application, it should be understood that the disclosed devices / computer equipment and methods can be implemented in other ways. For example, the device / computer equipment embodiments described above are only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0215] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0216] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0217] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0218] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for automatically switching high and low beams based on scene recognition, characterized in that: include: Acquire the light intensity of the road environment where the vehicle is located, identify the target object in the road environment using the first sensor combination, and obtain multiple target object recognition results and the target confidence corresponding to each target object recognition result; fusing a plurality of target confidences output by the first sensor combination to obtain a fused target confidence for the target object; Using the second sensor combination to identify lane lines in the road environment, and obtaining a plurality of lane line recognition results and a lane line confidence corresponding to each lane line recognition result; fusing the multiple lane line confidences output by the second sensor combination to obtain a fused lane line confidence; Determine the size of the target object from the multiple target object recognition results, determine the parameters of the lane line from the multiple lane line recognition results, and calculate the speed and relative position relationship of the target object according to the size of the target object and the parameters of the lane line; Constructing a scene data set according to the illumination intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed of the target object and the relative position relationship; The driving scene in which the current vehicle is located is identified based on the scene data set, and the corresponding high and low beam automatic switching control strategy is executed according to the identification result of the driving scene.

2. The method according to claim 1, characterized in that The first sensor combination includes a laser radar, a multi-function camera, and a millimeter wave radar, and the fusing of multiple target confidences output by the first sensor combination to obtain a fused target confidence for the target object includes: Obtaining a first target confidence level for identifying and outputting a target object using a laser radar, obtaining a second target confidence level for identifying and outputting a target object using a multi-function camera, and obtaining a third target confidence level for identifying and outputting a target object using a millimeter-wave radar; The first target confidence, the second target confidence and the third target confidence are weightedly calculated according to the preset weighting factors corresponding to different target confidences to obtain the fused target confidence.

3. The method according to claim 1, characterized in that The second sensor combination includes a laser radar and a multi-function camera, and the fusing of multiple lane line confidences output by the second sensor combination to obtain a fused lane line confidence includes: Obtaining a first lane line confidence value output by using a laser radar to identify a lane line in a road environment; Obtaining a second lane line confidence value output by recognizing a lane line in a road environment using a multifunctional camera; The first lane line confidence and the second lane line confidence are weightedly calculated according to the preset weighting factors corresponding to different lane line confidences to obtain the fused lane line confidence.

4. The method according to claim 1, characterized in that: Before determining the size of the target object from the plurality of target object recognition results, the method further includes: The fusion target confidence is compared with a preset fusion target confidence threshold. When the fusion target confidence exceeds the fusion target confidence threshold, the target object is determined as the actual target; the fusion lane line confidence is compared with a preset fusion lane line confidence threshold. When the fusion lane line confidence exceeds the fusion lane line confidence threshold, the lane line is determined as the actual lane line.

5. The method according to claim 1, characterized in that The determining the parameters of the lane line from the multiple lane line recognition results includes: Process the road image to extract pixel data containing lane line information, and collect point cloud data on the road surface to obtain discrete point data containing lane line information; The pixel point data and the discrete point data are processed separately to obtain a discrete point set of the lane line, and the discrete point set is merged and fitted to generate width and curvature parameters of the lane line; The curvature direction of the lane line is determined according to the curvature parameter, and it is determined whether the lane line is a straight lane.

6. The method according to claim 5, characterized in that The step of merging and fitting the discrete point set to generate lane line width and curvature parameters includes: According to the coordinate data set of the discrete points of the lane line, the discrete points are curve fitted by using the least square method to obtain the lane line equation; based on the lane line equation, the first-order derivative and the second-order derivative of the lane line are calculated, and the curvature and the curvature radius of the lane line are calculated based on the first-order derivative and the second-order derivative; Among them, the curvature is used to characterize the degree of curvature of the lane line. When the curvature is a positive value, it means that the lane is curved to the left; when the curvature is a negative value, it means that the lane is curved to the right; when the curvature is close to 0, it means that the lane line is a straight line.

7. The method according to claim 1, characterized in that The method of executing a corresponding high and low beam automatic switching control strategy according to the recognition result of the driving scene includes: When the driving scene is that the current vehicle is driving on a straight road or a curve, and the light intensity meets the preset high beam on condition, and when a target vehicle is identified in the opposite lane, the current vehicle is automatically controlled to turn off the high beam and switch to the low beam; When the driving scenario is that the current vehicle is driving on a straight road or a curve, the light intensity meets the preset high beam turn-on condition, and when a target vehicle is identified in an adjacent lane and it is determined that the current vehicle has an intention to overtake, the headlights of the current vehicle are automatically controlled to flash and switch to low beam.

8. A high and low beam automatic switching control device based on scene recognition, characterized in that: include: A first recognition module is configured to obtain the light intensity of the road environment where the vehicle is located, identify the target object in the road environment using the first sensor combination, and obtain multiple target object recognition results and the target confidence corresponding to each target object recognition result; a first fusion module configured to fuse a plurality of target confidences output by the first sensor combination to obtain a fused target confidence for the target object; A second recognition module is configured to recognize lane lines in the road environment using a second sensor combination to obtain a plurality of lane line recognition results and a lane line confidence corresponding to each lane line recognition result; A second fusion module is configured to fuse the multiple lane line confidences output by the second sensor combination to obtain a fused lane line confidence; a determination module, configured to determine the size of the target object from the multiple target object recognition results, determine the parameters of the lane line from the multiple lane line recognition results, and calculate the speed and relative position relationship of the target object according to the size of the target object and the parameters of the lane line; A construction module is configured to construct a scene data set according to the illumination intensity, the fused target confidence, the fused lane line confidence, the size of the target object, the parameters of the lane line, the speed of the target object and the relative position relationship; The control module is configured to identify the driving scene in which the current vehicle is located based on the scene data set, and execute a corresponding high and low beam automatic switching control strategy according to the recognition result of the driving scene.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Citation Information

Cited By

  • Lamp holder integrated controller system and energy-saving control method

    CN119389103A

  • Vehicle infotainment active response method and system used in vehicle low-speed scene

    CN121708560A

  • A method and system for active response of a vehicle machine in a low-speed scenario of a vehicle

    CN121708560B