A method, device, and electronic equipment for intelligent automatic headlight control

Abstract

This invention provides a method, apparatus, and electronic device for intelligent automatic headlight control. The method includes the following steps: acquiring an external environmental image and current illuminance using the vehicle's domain controller; when the current illuminance is less than or equal to a preset threshold, detecting a target object using a preset light source detection model to obtain target information; if the environmental image indicates the presence of other vehicles, detecting the environmental image using a preset deep learning detection model to obtain detection results, including whether a vehicle headlight source exists in the environmental image, and the distance and direction information between the vehicle headlight source and the vehicle when a vehicle headlight source exists; if the detection results and target information meet preset conditions, turning on the vehicle's headlights. This helps reduce the overall cost of the device and improve the accuracy of headlight control.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence technology, specifically relating to a method, device, and electronic equipment for automatic control of intelligent headlights. Background Technology

[0002] During driving, the headlights may need to be adjusted under different driving conditions. Currently, the mainstream automatic headlight control system detects the external light source environment by adding corresponding sensors outside the vehicle.

[0003] Currently, patent CN207274535U discloses a control system for intelligent headlights in automobiles. This solution uses an independent hardware control module, intelligent headlights, and photosensitive sensors to receive light data and control the turning of the intelligent headlights on and off. It also uses information such as steering wheel angle and vehicle speed for auxiliary headlight control. This solution, using a hardware module and independent controller to control the intelligent headlights, increases costs and does not fully consider the future development direction of intelligent vehicle domain controllers. Patent CN114162037A discloses a vehicle intelligent headlight system based on a forward-looking camera. This solution uses the acquired forward-looking camera image as input to the headlight system and integrates vehicle status information to control the switching and direction of the headlights. This solution only uses image and vehicle body information for judgment, limiting the external vehicle information it can acquire and utilize, and is prone to misjudgments and invalid detections.

[0004] It is evident that the mainstream solutions currently used for controlling headlights in the market have increased additional costs, resulting in high costs or low control precision, which is not conducive to precise control of headlight switches. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method, device and electronic device for intelligent headlight automatic control, so as to improve the problems of high cost and low control accuracy of the current mainstream solutions.

[0006] The present invention solves the above-mentioned technical problems through the following technical means:

[0007] In a first aspect, embodiments of this application provide a method for intelligent headlight control, including:

[0008] Use the vehicle's domain controller to obtain an image of the external environment and the current illuminance.

[0009] When the current illuminance is less than or equal to a preset threshold, the target object is detected by a preset light source detection model to obtain target information;

[0010] If the environmental image indicates the presence of other vehicles, the environmental image is detected using a preset deep learning detection model to obtain detection results. The detection results include whether there are vehicle headlight sources in the environmental image, and the distance and direction information between the vehicle headlight sources and the vehicle when the vehicle headlight sources are present.

[0011] If the detection result and the target information meet the preset conditions, then turn on the vehicle's headlights.

[0012] In conjunction with the first aspect, in some optional implementations, the method further includes: acquiring an environmental image of the vehicle's exterior and the current illuminance using the vehicle's domain controller; and, when the current illuminance is less than or equal to a preset threshold, detecting the target object using a preset light source detection model to obtain target information.

[0013] When the current illuminance is greater than the preset threshold, the headlights of this vehicle are controlled to be turned off.

[0014] In conjunction with the first aspect, in some optional implementations, the target object is detected using a preset light source detection model to obtain target information, including:

[0015] The detection object is acquired using a forward-facing camera;

[0016] The light source information is obtained by detecting the object using a preset light source detection model.

[0017] The light source information is classified and filtered to obtain target information. The filtering of the light source information includes deleting invalid light sources in the light source detection box through the light source detection model. The invalid light sources include the background, sky, street lights in the image, and light sources in the environment image whose output light source confidence is less than the confidence threshold.

[0018] In conjunction with the first aspect, in some optional implementations, the light source information includes: light source type, lateral and longitudinal distances between the light source and the vehicle in the VCS coordinate system, and light source confidence level, wherein the light source type includes headlights, taillights, and streetlights.

[0019] In conjunction with the first aspect, in some optional implementations, the target information includes:

[0020] The detection frame of the light source, the light source of the type of headlight or taillight, the light source with a confidence level greater than or equal to the confidence level threshold, and the lateral and longitudinal distances between the light source and the vehicle in the VCS coordinate system.

[0021] In conjunction with the first aspect, in some alternative embodiments, the method further includes, before turning on the vehicle's headlights:

[0022] The deep learning detection model is used to fuse the light source detection results of the front and rear of the vehicle. If two light source detection boxes of the same type have overlapping areas, the two light source detection results are fused into one light source detection result; otherwise, they are two light source detection results.

[0023] The processed light source detection results are combined with the target information. Light sources with a confidence threshold greater than or equal to the target light source are selected as target light sources. The lateral distance and longitudinal distance between the vehicle and the target light source in the VCS coordinate system are determined to meet the preset conditions. The preset conditions are determined to be met when the lateral distance is within a first preset distance and the longitudinal distance is within a second preset distance.

[0024] In conjunction with the first aspect, in some alternative implementations, the preset threshold is set to 10 lux.

[0025] Secondly, embodiments of this application also provide a device for automatic control of intelligent headlights, comprising:

[0026] The first information acquisition unit uses the vehicle's domain controller to acquire images of the external environment and the current illuminance of the vehicle.

[0027] The first execution unit, when the current illuminance is less than or equal to a preset threshold, performs detection on the detection object using a preset light source detection model to obtain target information;

[0028] The second information acquisition unit, if the environmental image indicates the presence of other vehicles, then detects the environmental image using a preset deep learning detection model to obtain a detection result. The detection result includes whether there is a vehicle light source in the environmental image, and the distance and direction information between the vehicle light source and the vehicle when the vehicle light source is present.

[0029] The second execution unit turns on the vehicle's headlights if the detection result and the target information meet preset conditions.

[0030] Thirdly, embodiments of this application also provide an electronic device, which includes a processor and a memory coupled to each other, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the electronic device performs the above-described method.

[0031] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the methods described above.

[0032] The present invention has the following advantages:

[0033] The electronic device uses the image processing of the ISP unit in the domain controller to acquire the external environment image and current illuminance of the vehicle. If the current illuminance is greater than a preset threshold, the electronic device controls the vehicle's headlights to be off. If the current illuminance is less than or equal to the preset threshold, the electronic device uses a preset light source detection model to detect the target object, acquire light source information, classify and filter the light source information to obtain target information. If the environment image indicates the presence of other vehicles, a preset deep learning detection model is used to detect the environment image, obtain detection results, and the electronic device combines the detection results with the target information to determine whether it meets preset conditions. If the detection results and target information meet the preset conditions, the vehicle's headlights are turned on. By using existing cameras and the domain controller to detect the surrounding environment, using deep learning detection models and light source detection models to detect the required information and combine them, and finally using the electronic device to judge and control the headlights to turn on and off, there is no need to add a separate controller to achieve the effect of controlling the headlights. This directly uses the existing hardware system, which helps to reduce the overall hardware cost of the system. In addition, this application does not only use images and vehicle body information for judgment, but also uses external environmental images, illumination, and the results of the detection object obtained by the forward-looking camera to continuously cross-combine and gradually obtain the final result, and uses this as the basis for judgment to improve the reliability of the judgment, thereby achieving high control accuracy and improving the problems of high cost and low control accuracy of existing mainstream solutions. Attached Figure Description

[0034] The present invention can be further illustrated by the non-limiting embodiments given in the accompanying drawings:

[0035] Figure 1 This is a flowchart illustrating a method for automatic control of intelligent headlights provided by an embodiment of the present invention;

[0036] Figure 2 This is a block diagram of an intelligent headlight automatic control device provided by an embodiment of the present invention.

[0037] Icons: 200 - First information acquisition unit; 210 - First execution unit; 220 - Second information acquisition unit; 230 - Second execution unit. Detailed Implementation

[0038] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that similar or identical parts are referred to by the same reference numerals in the drawings or description. Implementations not shown or described in the drawings are forms known to those skilled in the art. In the description of this application, terms such as "first" and "second" are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0039] This application provides an electronic device. The electronic device includes a processor and a memory coupled together. The memory stores a computer program. When the computer program is executed by the processor, the electronic device is able to perform the corresponding steps in the following method for automatic control of intelligent headlights.

[0040] Electronic devices can be, but are not limited to, personal computers, servers, etc. The processor and memory are electrically connected; the processor executes computer programs stored in the memory. Since both the processor and memory are existing hardware modules, they will not be elaborated upon here.

[0041] like Figure 1 As shown, this application also provides a method for automatic control of intelligent headlights, which can be applied to the aforementioned electronic device, and the electronic device executes or implements the steps of the method. The method for automatic control of intelligent headlights may include the following steps:

[0042] Step 100: Use the vehicle's domain controller to obtain the environmental image and illuminance information outside the vehicle;

[0043] Step 110: When the current illuminance is less than or equal to a preset threshold, the target object is detected by a preset light source detection model to obtain target information;

[0044] Step 120: If the environmental image indicates the presence of other vehicles, the environmental image is detected using a preset deep learning detection model to obtain detection results. The detection results include whether there are vehicle headlight sources in the environmental image, and the distance and direction information between the vehicle headlight sources and the vehicle when vehicle headlight sources are present.

[0045] Step 130: If the detection results and target information meet the preset conditions, turn on the headlights of this vehicle.

[0046] The steps of the intelligent headlight automatic control method will be explained in detail below:

[0047] In step 100, the electronic device uses the image processing of the ISP (Internet Service Provider) unit in the domain controller to obtain an image of the external environment of the vehicle and the current illuminance.

[0048] In this embodiment, the current illuminance refers to the intensity of light in the external environment where the vehicle is located, as detected by the ISP. Specifically, the illuminance is calculated and processed by the ISP using a BAYER format environmental image; subsequently, the illuminance is output externally through the interface provided by the ISP module.

[0049] In step 110, when the current illuminance is less than or equal to a preset threshold, the electronic device detects the target object using a preset light source detection model to obtain target information.

[0050] In this embodiment, the preset threshold is set to 10 lux. A first judgment result is obtained by comparing the current illuminance with the preset threshold. This first judgment result includes information that the current illuminance is greater than or less than the preset threshold. That is, if the current illuminance is greater than the preset threshold, the headlights are turned off; if the current illuminance is less than or equal to the preset threshold, the light source detection model is activated to detect the target object. In this embodiment, lux is a unit of illuminance intensity.

[0051] In this embodiment, the detection objects include other light sources that enter the field of view of the camera, excluding the vehicle itself. The light source detection model detects these light sources to obtain light source information.

[0052] In this embodiment, the electronic device acquires the detection object through a front-view camera, then detects the object using a preset light source detection model to obtain light source information. This light source information is then classified and filtered to obtain target information. Filtering the light source information includes deleting invalid light sources from the light source detection frame using the light source detection model. Invalid light sources include the background, sky, streetlights, and light sources in the environmental image whose output confidence level is less than a confidence threshold. The confidence threshold can be flexibly set according to actual conditions, for example, 10%. Light sources with a confidence level less than the threshold are less likely to actually be "light sources" and need to be deleted.

[0053] In this embodiment, the light source information includes the light source type, the lateral and longitudinal distances of the light source in the VCS (Vertical Coordinate System) coordinate system, and the light source confidence level; wherein, the light source type includes headlights, taillights, and streetlights.

[0054] In this embodiment, the target information includes the detection frame of the light source, the light source of the type of headlight or taillight, the light source with a confidence level greater than or equal to the confidence level threshold, and the horizontal and vertical distance of the light source in the VCS coordinate system.

[0055] In step 120, if the environmental image indicates the presence of other vehicles, the electronic device detects the environmental image using a preset deep learning detection model to obtain detection results. The detection results include whether there are vehicle headlight sources in the environmental image, and the distance and direction information between the vehicle headlight sources and the vehicle when vehicle headlight sources are present.

[0056] In this embodiment, the electronic device determines whether other vehicles exist based on the input environmental image. It determines the presence of other vehicles by counting the number of rear-view frames of each vehicle. For example, if the number of rear-view frames is greater than 3, then other vehicles exist, and the next step is performed; if the number of rear-view frames is less than or equal to 3, then no other vehicles exist, and the headlights are turned off.

[0057] In this embodiment, the deep learning detection model can be a CNN model or other models. While a single deep learning detection model is existing technology, in this embodiment, the deep learning detection model is combined with other structures to infer the specific light source conditions around the vehicle through light source detection.

[0058] In this embodiment, if other vehicles are in motion, the method may further include the following steps before step 130:

[0059] The detection results of light sources at the front and rear of the vehicle are fused using a deep learning detection model. If two light source detection boxes of the same type have overlapping areas, the two light source detection results are fused into one light source detection result; otherwise, they are two light source detection results.

[0060] The processed light source detection results are combined with the target information. Light sources with a confidence threshold greater than or equal to the target light source are used as target light sources. It is then determined whether the lateral distance and longitudinal distance between the vehicle and the target light source in the VCS coordinate system meet the preset conditions. Specifically, if the lateral distance is within the first preset distance and the longitudinal distance is within the second preset distance, it is determined that the preset conditions are met.

[0061] In this embodiment, the processed light source detection results are combined with target information. That is, after comprehensively organizing the light source detection results and target information, the decision to turn on the headlights is based on whether the lateral and longitudinal distances between the vehicle and the target light source in the VCS coordinate system meet preset conditions. The first and second preset distances can be flexibly determined according to actual circumstances.

[0062] In step 130, if the detection result and target information meet the preset conditions, the headlights of the vehicle are turned on.

[0063] For example, taking the vehicle as the center point, if the VCS coordinate system displays light source information within a lateral distance of 30m and a longitudinal distance of 220m, then the detection result and target information meet the preset conditions, and the next step is to turn on the vehicle's headlights; if the VCS coordinate system displays light source information where the lateral distance exceeds 30m or the longitudinal distance exceeds 220m, then the detection result and target information do not meet the preset conditions, and the headlights are turned off.

[0064] This application also provides a device for automatic control of intelligent headlights, which includes at least one software function module stored in a storage module or embedded in an operating system in the form of software or firmware. A processor is used to execute the executable modules stored in the storage module, such as the software function modules and computer programs included in the device for automatic control of intelligent headlights.

[0065] like Figure 2 As shown, the intelligent headlight automatic control device includes a first information acquisition unit 200, a first execution unit 210, a second information acquisition unit 220, and a second execution unit 230. The functions of each unit are as follows:

[0066] The first information acquisition unit 200 uses the vehicle's domain controller to acquire an image of the external environment and the current illuminance of the vehicle.

[0067] The first execution unit 210, when the current illuminance is less than or equal to a preset threshold, performs detection on the detection object using a preset light source detection model to obtain target information;

[0068] The second information acquisition unit 220, if the environmental image indicates the presence of other vehicles, then detects the environmental image using a preset deep learning detection model to obtain detection results. The detection results include whether there are vehicle light sources in the environmental image, and the distance and direction information between the vehicle light sources and the vehicle when vehicle light sources are present.

[0069] If the detection result and target information meet the preset conditions, the second execution unit 230 turns on the headlights of the vehicle.

[0070] Optionally, the first execution unit 210 can be used to: set a threshold for illuminance, including: setting a preset threshold to 10 lux.

[0071] Optionally, between the first information acquisition unit 200 and the first execution unit 210, a third execution unit 240 is also included. The third execution unit 240 can be used to control the headlights of the vehicle to be turned off when the current illuminance is greater than a preset threshold.

[0072] Optionally, the first execution unit 210 can be used to: acquire a detection object through a forward-looking camera; detect the detection object using a preset light source detection model to obtain light source information; classify and filter the light source information to obtain target information, wherein filtering the light source information includes deleting invalid light sources in the light source detection frame using the light source detection model. Invalid light sources include the background, sky, streetlights in the image, and light sources in the environmental image whose output light source confidence is less than the confidence threshold.

[0073] The light source information includes the light source type, the lateral and longitudinal distances between the light source and the vehicle in the VCS coordinate system, and the light source confidence level. The light source type includes headlights, taillights, and streetlights. The target information includes the detection frame of the light source, the light source type of headlights or taillights, the light source confidence level greater than or equal to the confidence level threshold, and the lateral and longitudinal distances between the light source and the vehicle in the VCS coordinate system.

[0074] Optionally, the second execution unit 230 can be used to: fuse the light source detection results of the front and rear of the vehicle using a deep learning detection model; if two light source detection boxes of the same type have overlapping areas, then the two light source detection results are fused into one light source detection result; otherwise, they are two light source detection results.

[0075] The processed light source detection results are combined with the target information. Light sources with a confidence threshold greater than or equal to the target light source are used as target light sources. It is then determined whether the lateral distance and longitudinal distance between the vehicle and the target light source in the VCS coordinate system meet the preset conditions. Specifically, if the lateral distance is within the first preset distance and the longitudinal distance is within the second preset distance, it is determined that the preset conditions are met.

[0076] This application also provides a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the intelligent headlight automatic control method described in the above embodiments.

[0077] Based on the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by hardware or by using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, control device, or network device, etc.) to execute the methods described in the various implementation scenarios of this application.

[0078] In summary, this application provides a method, apparatus, and electronic device for intelligent automatic headlight control. In this solution, the electronic device uses image processing in the ISP unit of the domain controller to acquire an external environmental image and current illuminance. If the current illuminance is greater than a preset threshold, the electronic device controls the vehicle's headlights to be off. If the current illuminance is less than or equal to the preset threshold, the electronic device uses a preset light source detection model to detect the target object, acquires light source information, and classifies and filters the light source information to obtain target information. If the environmental image indicates the presence of other vehicles, a preset deep learning detection model is used to detect the environmental image, obtain detection results, and the electronic device combines the detection results with the target information to determine whether they meet preset conditions. If the detection results and target information meet the preset conditions, the vehicle's headlights are turned on. This application uses existing cameras and domain controllers to detect the surrounding environment, employs deep learning detection models and light source detection models to detect and combine the required information, and finally uses electronic devices to judge and control the headlights to turn on and off. This eliminates the need for an additional independent controller to control the headlights, directly using existing devices and reducing the overall cost of the device. Furthermore, this application does not only rely on images and vehicle body information for judgment, but also uses external environmental images, illuminance, and the results of detection objects obtained by the forward-looking camera to continuously cross-combine and gradually derive the final result, which is then used as the basis for judgment. This results in a lower overall cost and higher control accuracy, improving upon the problems of high cost and low control accuracy in existing mainstream solutions.

[0079] In the embodiments provided in this application, it should be understood that the disclosed apparatus, devices, and methods can also be implemented in other ways. The apparatus, devices, and methods embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which includes one or more executable instructions for implementing a specified logical function. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions. Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0080] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method of intelligent headlight control, characterized in that The method comprises the following steps: acquiring, by a domain controller of the vehicle, an environment image outside the vehicle and a current illumination; when the current illumination is less than or equal to a preset threshold, detecting a detection object by a preset light source detection model to obtain target information; if the environment image indicates that there are other vehicles, detecting the environment image by a preset deep learning detection model to obtain a detection result, the detection result comprising whether there is a vehicle light source in the environment image, and the distance and direction information of the vehicle light source from the vehicle if the vehicle light source exists; and if the detection result and the target information meet a preset condition, turning on the high beam of the vehicle; wherein the target information is obtained by detecting the detection object by the preset light source detection model, comprising: acquiring the detection object by a front-view camera; detecting the detection object by the preset light source detection model to obtain light source information; classifying and filtering the light source information to obtain the target information, wherein the filtering of the light source information comprises deleting invalid light sources in a light source detection frame by the light source detection model, the invalid light sources comprising the background, sky and street lamp on the image, and light sources with a light source confidence less than a confidence threshold in the environment image; the target information comprising a detection frame of the light source, a headlight or taillight type light source, a light source with a light source confidence greater than or equal to the confidence threshold, and the lateral and longitudinal distances of the light source from the vehicle in a VCS coordinate system; before turning on the high beam of the vehicle, the method further comprises: fusing the light source detection results of the head and tail of the vehicle by the deep learning detection model, if two light source detection frames of the same type have an overlapping area, fusing the two light source detection results into one light source detection result, otherwise, keeping the two light source detection results; combining the processed light source detection results with the target information, taking the light source with a light source confidence greater than or equal to the confidence threshold as a target light source, and determining whether the lateral and longitudinal distances of the vehicle and the target light source in the VCS coordinate system meet the preset condition, wherein the lateral distance is within a first preset distance and the longitudinal distance is within a second preset distance, and it is determined that the preset condition is met.

2. The method of claim 1, wherein, In the step of acquiring, by a domain controller of the vehicle, an environment image outside the vehicle and a current illumination, and the step of detecting a detection object by a preset light source detection model to obtain target information when the current illumination is less than or equal to a preset threshold, the method further comprises: when the current illumination is greater than the preset threshold, controlling the high beam of the vehicle to be in an off state.

3. The method of claim 1, wherein, The light source information comprises the light source type, the lateral and longitudinal distances of the light source from the vehicle in the VCS coordinate system, and the light source confidence, wherein the light source type comprises the headlight, the taillight and the street lamp.

4. The method of claim 1, wherein, The preset threshold is set to 10 lux.

5. A device for automatic control of intelligent headlamps, characterized in that The device comprises a first acquisition information unit for acquiring, by a domain controller of the vehicle, an environment image outside the vehicle and a current illumination; ​ The first execution unit detects a detection object by using a preset light source detection model to obtain target information when the current illumination is less than or equal to a preset threshold value. The second acquisition information unit detects the environment image by using a preset deep learning detection model to obtain a detection result if the environment image indicates that there is a remaining vehicle, wherein the detection result includes whether there is a vehicle light source in the environment image, and distance and direction information of the vehicle light source from the vehicle if the vehicle light source exists. The second execution unit turns on a high beam of the vehicle if the detection result and the target information meet a preset condition.

6. An electronic device, comprising: The electronic device includes a processor and a memory coupled to each other, and the memory stores a computer program. When the computer program is executed by the processor, the electronic device executes the method of any one of claims 1-4.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores a computer program. When the computer program runs on a computer, the computer executes the method of any one of claims 1-4.

Images