Vehicle lamp control method and device and vehicle

Abstract

The invention provides a vehicle lamp control method and device and a vehicle, and relates to the technical field of automatic driving. The vehicle lamp control method comprises the steps of obtaining an environment image; obtaining a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of an identified object in the environment image; determining a target distance according to the plurality of initial distances and the confidence of each initial distance; and outputting a light control signal according to the target distance, wherein the light control signal is used for controlling the irradiation range of the vehicle lamp.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and more particularly to a vehicle light control method, device, and vehicle. Background Art

[0002] The headlight control in related technologies usually adopts a preset illumination mode or controls the illumination range of the headlights based on the ambient light intensity. The above control method cannot accurately reflect the complexity and dynamic changes of the vehicle's surrounding environment, resulting in inaccurate setting of the headlight illumination range, affecting the safety of vehicle driving. Summary of the Invention

[0003] In view of this, the present disclosure provides a vehicle light control method, device, and vehicle.

[0004] A first aspect of the present disclosure provides a vehicle light control method, comprising:

[0005] Get the environment image;

[0006] Obtaining a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image;

[0007] determining a target distance according to the multiple initial distances and a confidence level of each initial distance;

[0008] A light control signal is output according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

[0009] According to an embodiment of the present disclosure, the confidence level of the initial distance is obtained by predicting the target model based on the position of the identified object in the environment image and / or the brightness of the identified object;

[0010] The determining the target distance according to the multiple initial distances and the confidence level of each initial distance includes:

[0011] determining at least one candidate distance from the plurality of initial distances, the candidate distance being an initial distance having a confidence greater than a confidence threshold;

[0012] The candidate distance with the largest distance among the at least one candidate distance is used as the target distance.

[0013] According to an embodiment of the present disclosure, outputting a light control signal according to the target distance includes:

[0014] Get the vehicle driving speed;

[0015] A light control signal is output based on the vehicle driving speed and the target distance.

[0016] According to an embodiment of the present disclosure, the outputting the light control signal according to the target distance comprises:

[0017] obtaining speed limit information, the speed limit information comprising at least one of road condition information of a current driving road of the vehicle, regulation information, and preset information input by a user;

[0018] determining a minimum driving speed based on the speed limit information;

[0019] outputting the light control signal based on the minimum driving speed and the target distance.

[0020] According to an embodiment of the present disclosure, the method comprises:

[0021] obtaining driving information, the driving information being determined based on a current motion trajectory of the vehicle and a driving speed of the vehicle;

[0022] determining a target region based on the driving information;

[0023] selecting at least one candidate distance from initial distances corresponding to identified objects located in the target region;

[0024] determining the target distance according to each of the candidate distances and a corresponding confidence level.

[0025] According to an embodiment of the present disclosure, the method comprises:

[0026] obtaining a driving speed of the vehicle and an ambient brightness;

[0027] determining a first perception range according to a change of the driving speed of the vehicle and the ambient brightness, the first perception range representing a region required to be identified by the vehicle under the ambient brightness;

[0028] determining a second perception range based on the plurality of initial distances, the second perception range representing a region identifiable by the target model under the ambient brightness;

[0029] outputting a light control signal based on the first perception range and the second perception range.

[0030] According to an embodiment of the present disclosure, the method comprises:

[0031] determining a target perception range based on a difference between the first perception range and the second perception range, the target perception range representing a perception capability of the target model under the driving speed of the vehicle and the ambient brightness.

[0032] According to an embodiment of the present disclosure, the determining the first perception range based on the driving speed of the vehicle comprises:

[0033] determining a first longitudinal distance required to be identified for the vehicle under the ambient brightness according to the vehicle driving speed;

[0034] Determining a first lateral distance of the vehicle that needs to be identified under the ambient brightness according to the road on which the vehicle is located;

[0035] A first perception range is determined based on the first longitudinal distance and the first lateral distance.

[0036] A second aspect of the present disclosure provides a vehicle light control device, comprising:

[0037] Acquisition module, used to obtain environmental images;

[0038] A first determination module is configured to obtain a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image;

[0039] a second determining module, configured to determine a target distance based on the multiple initial distances and a confidence level of each initial distance;

[0040] The output module outputs a light control signal according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

[0041] A third aspect of the present disclosure provides a vehicle, comprising:

[0042] Vehicle body;

[0043] A headlight, arranged on the vehicle body;

[0044] A collection device, provided on the vehicle body, for acquiring an environmental image;

[0045] a first controller, disposed in the vehicle body, configured to obtain a plurality of initial distances determined by a target model based on the environment image, the initial distances representing corresponding distances of recognized objects in the environment image; determine a target distance based on the plurality of initial distances and a confidence level of each initial distance; and output a light control signal based on the target distance;

[0046] The second controller is provided in the vehicle body and is used for receiving the light control signal and controlling the illumination range of the vehicle light according to the light control signal.

[0047] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0049] Figure 1 A flow chart of a vehicle light control method provided by an embodiment of the present disclosure is schematically shown;

[0050] Figure 2 A schematic diagram of a scenario of a vehicle light control method provided by an embodiment of the present disclosure;

[0051] Figure 3 The following schematically shows an architecture block diagram of a vehicle light control system according to an embodiment of the present disclosure;

[0052] Figure 4 The following schematically shows a signaling interaction diagram under a vehicle light control system provided in accordance with an embodiment of the present disclosure;

[0053] Figure 5 The following schematically shows a structural block diagram of a vehicle light control device according to an embodiment of the present disclosure;

[0054] Figure 6 The figure schematically shows a structural block diagram of a vehicle according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0055] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that the above description is merely exemplary and is not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without the above-mentioned specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0056] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise," "include," etc. used herein indicate the presence of the features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0057] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0058] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0059] In the embodiments of this disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, and storage of all data involved (including, but not limited to, user personal information) comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures have been taken to prevent unauthorized access to user personal information data and to safeguard the security of user personal information, network security, and national security.

[0060] The present disclosure provides a vehicle light control method, including:

[0061] Get the environment image;

[0062] Obtaining a plurality of initial distances determined by the target model according to the environment image, where the initial distances represent corresponding distances of the identified objects in the environment image;

[0063] determining a target distance based on a plurality of initial distances and a confidence level of each initial distance;

[0064] The light control signal is output according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

[0065] By adopting the disclosed embodiments, an environmental image is acquired and analyzed and identified using a target model, enabling direct acquisition of multiple initial distances for identified objects in the environmental image, thereby enabling real-time perception of the vehicle's surroundings. The target model provides a confidence assessment for each initial distance, and the target distance is determined based on the multiple initial distances and corresponding confidence levels. This allows the target distance determination process to consider the reliability of the recognition results, improving the accuracy of distance judgment. Based on the accurately determined target distance, a light control signal is output to control the headlight illumination range, enabling the headlight illumination range to be controlled based on the actual position and distance of the identified object in the actual driving environment. This resolves the issue of inaccurate headlight illumination range control in related technologies and improves vehicle safety in complex and dynamic driving environments.

[0066] Figure 1 The figure schematically shows a flow chart of a vehicle light control method provided by an embodiment of the present disclosure.

[0067] like Figure 1As shown, the vehicle lamp control method specifically includes operation S101 to operation S104.

[0068] Operation S101, acquiring an environment image;

[0069] Operation S102, obtaining a plurality of initial distances determined by a target model according to the environment image, the initial distances representing corresponding distances of the recognized objects in the environment image;

[0070] Operation S103, determining a target distance according to the plurality of initial distances and the confidence of each initial distance;

[0071] Operation S104, outputting a light control signal according to the target distance, the light control signal being used for controlling the irradiation range of the vehicle lamp.

[0072] In operation S101, the environment image refers to real-time image data of the road environment in front of and around the vehicle, which can be understood in the embodiment of the present disclosure as visual information collected by the vehicle-mounted camera for automatic driving perception analysis, for subsequent target model to identify and distance calculate the objects in the environment around the vehicle.

[0073] For example, the environment image includes but is not limited to: vehicle front road image, vehicle side road image, road image containing other vehicles, road image containing pedestrians or obstacles, road image under different light conditions, etc.

[0074] In a feasible implementation, the environment image of the road in front of the vehicle can be acquired in real time by a single camera installed in front of the vehicle, and the camera continuously acquires the environment image at a preset frame rate, and each frame of the environment image contains the visual information of the objects such as the road conditions, other vehicles, pedestrians, traffic signs, etc. in front of the vehicle at the current time.

[0075] In another feasible implementation, the environment image is acquired by a combination of multiple cameras, including a front camera, a left camera and a right camera, which respectively acquire the road environment images in front of, left side and right side of the vehicle, and the images acquired by the multiple cameras are fused to form an environment image covering a wider visual angle range around the vehicle, so as to obtain more comprehensive environment perception information.

[0076] It should be noted that the acquisition of the environment image is not limited to visible light image, but also includes infrared image, night vision image, etc. to adapt to the environment perception demand under different light conditions and ensure that effective environment image can be acquired under various driving conditions for subsequent target recognition and distance analysis.

[0077] In operation S102, the target model refers to an algorithm model for environmental perception and target detection built based on deep learning or machine learning technology. In the embodiment of the present disclosure, it can be understood as a neural network model that can identify and classify various objects in the environmental image and estimate their distance from the vehicle, and is used to achieve accurate measurement of the position and distance of the identified objects in the vehicle's surrounding environment.

[0078] Exemplarily, the target model includes but is not limited to a convolutional neural network model, a recurrent neural network model, a transformer model, a target detection model, a multi-task learning model, etc.

[0079] It should be noted that the target model can be any target detection and distance estimation model available in the prior art. Target models already have mature technical solutions in the field of autonomous driving. Therefore, we will not elaborate on the specific network structure, training dataset, training methods, and other technical details of the target model. The present disclosure focuses on how to use the output of the target model to control vehicle lights.

[0080] Similarly, the target model can be used to determine the identified objects in the environment map. The identified objects refer to the various identified road environment elements. In the embodiment of the present disclosure, they can be understood as various objects or entities that need to be perceived and avoided during vehicle driving, and are used for vehicle path planning, obstacle avoidance decisions and environmental perception for safe driving.

[0081] Exemplarily, the identified objects include, but are not limited to, motor vehicles, non-motor vehicles, pedestrians, animals, static obstacles, road infrastructure, and road boundary markings.

[0082] Furthermore, the identified objects can also be classified according to their positional relationship relative to the vehicle, including front objects, side objects, rear objects, as well as static objects and dynamic objects classified according to their motion state.

[0083] Similarly, the initial distance refers to the spatial distance value between the vehicle and each identified object in the environmental image calculated by the target model. In the embodiment of the present disclosure, it can be understood as the straight-line distance or the distance along the driving path from the vehicle to the identified object determined by image analysis technology, which is used to evaluate the degree of impact and urgency of the identified object on the vehicle driving safety.

[0084] It should be noted that the same identified object may correspond to multiple initial distance values, because the target model may repeatedly detect the same object in multiple consecutive frames of images, or produce multiple detection results for the same object in a single frame of image, and each detection result will output a corresponding initial distance value.

[0085] In operation S103, the confidence of the initial distance refers to a quantitative indicator of the accuracy of each initial distance value output by the target model. In the embodiment of the present disclosure, it can be understood as a numerical parameter that reflects the reliability of the target model's distance measurement results for the identified object, and is used to evaluate and screen the most reliable distance measurement results among multiple initial distances.

[0086] Exemplarily, the confidence of the initial distance can be a value between 0 and 1, where a value close to 1 indicates that the target model has a higher confidence in the distance measurement result, and a value close to 0 indicates that the target model has a lower confidence in the distance measurement result.

[0087] Similarly, the target distance refers to the key distance parameter used for headlight control determined after confidence screening and analysis of multiple initial distances. In the disclosed embodiment, it can be understood as a characteristic distance value that reflects the vehicle's perception capability state in the current environment, and is used as the basis for controlling the headlight illumination range.

[0088] In one feasible implementation, a confidence threshold can be set, and then initial distances with confidence levels greater than the threshold can be selected from multiple initial distances as candidate distances. Finally, the candidate distance with the highest value among all candidate distances is selected as the target distance. This approach ensures that the selected target distance is both sufficiently reliable and reflects the vehicle's current maximum effective sensing capability.

[0089] Another feasible implementation involves performing a weighted average calculation of multiple initial distances based on their confidence levels. Initial distances with higher confidence levels are weighted more heavily in the average calculation, and the resulting weighted average is used as the target distance. This approach comprehensively considers all reliable distance information, avoids the potential for accidental errors associated with selecting a single distance value, and improves the stability and accuracy of target distance determination.

[0090] It should be noted that the confidence calculation can be based on a variety of technical parameters within the target model, including but not limited to factors such as the clarity of the identified object in the image, the stability of the bounding box, and the similarity of feature matching. Different types of target models may use different confidence calculation methods, and the embodiments of this disclosure do not limit the specific confidence calculation algorithm.

[0091] In operation S104, the light control signal refers to a control instruction generated based on the target distance to guide the vehicle lighting system to adjust the illumination parameters. In the embodiment of the present disclosure, it can be understood as a digital signal or analog signal containing control parameters such as the vehicle lighting range, illumination angle, and light intensity, which is used to drive the vehicle lighting actuator to achieve precise adjustment of the illumination range to ensure that the vehicle lighting effect matches the actual perception needs of the vehicle.

[0092] For example, the lighting control signal can be a digital control instruction that includes an illumination distance parameter, such as a command signal that reads "Set illumination distance to 150 meters." It can also be a composite signal that includes multiple control dimensions, such as a combination of control parameters for illumination distance, horizontal illumination angle, vertical illumination angle, and light intensity. The lighting control signal can also be a pulse-width modulation signal, a controller area network bus communication protocol signal, or a dedicated vehicle light control protocol signal.

[0093] In one feasible implementation, the target distance can be compared with a preset illumination distance threshold. When the target distance is less than the illumination distance threshold, a light control signal is generated to increase the illumination range, instructing the lighting system to expand the illumination range to compensate for insufficient perception. When the target distance is greater than or equal to the illumination distance threshold, a light control signal is generated to maintain the current illumination range, maintaining the current illumination state of the lights. This threshold comparison method enables dynamic lighting control based on perception.

[0094] In another feasible implementation, a continuous mapping relationship between target distance and illumination range can be established. The corresponding illumination distance can be calculated proportionally based on the target distance value, generating a lighting control signal containing precise illumination distance parameters. For example, when the target distance is 100 meters, the illumination distance is set to 120 meters, and when the target distance is 80 meters, the illumination distance is set to 150 meters. The conversion from target distance to illumination distance is achieved through a linear or nonlinear function relationship, and the corresponding lighting control signal is generated.

[0095] It's important to note that the generation of lighting control signals can also take into account factors such as the vehicle's current driving state, road conditions, and regulatory requirements, allowing for correction and optimization of initial control parameters calculated based on target distance. Furthermore, the output frequency and update period of the lighting control signals can be adjusted based on actual application requirements to balance control accuracy and system stability.

[0096] By adopting the disclosed embodiments, an environmental image is acquired and analyzed and identified using a target model, enabling direct acquisition of multiple initial distances for identified objects in the environmental image, thereby enabling real-time perception of the vehicle's surroundings. The target model provides a confidence assessment for each initial distance, and the target distance is determined based on the multiple initial distances and corresponding confidence levels. This allows the target distance determination process to consider the reliability of the recognition results, improving the accuracy of distance judgment. Based on the accurately determined target distance, a light control signal is output to control the headlight illumination range, enabling the headlight illumination range to be controlled based on the actual position and distance of the identified object in the actual driving environment. This resolves the issue of inaccurate headlight illumination range control in related technologies and improves vehicle safety in complex and dynamic driving environments.

[0097] In actual vehicle driving, the analysis of the target model on the environment image often produces a large number of detection results, which contain both high-quality reliable recognition and low-quality recognition results caused by poor light conditions, target occlusion, image blur and other factors. If all the initial distances are directly used for vehicle lamp control, the unreliable distance information will cause inaccurate control of the vehicle lamp illumination range, thereby affecting the driving safety of the vehicle.

[0098] To solve the above problems, on the basis of the above embodiment, as an optional embodiment, the confidence of the initial distance is predicted by the target model based on the position of the recognized object in the environment image and / or the brightness of the recognized object. On this basis, operation S103 can further include the following operations:

[0099] Operation S201, determining at least one candidate distance from the plurality of initial distances, the candidate distance being an initial distance with a confidence greater than a confidence threshold;

[0100] Operation S202, in the at least one candidate distance, the candidate distance with the largest distance is taken as the target distance.

[0101] In the embodiments of the present disclosure, the target model can predict the confidence of the initial distance through different implementations to adapt to different driving scenarios.

[0102] In a feasible implementation, the target model only predicts the confidence based on the position features of the recognized object in the environment image. This method is suitable for daytime driving scenarios with relatively stable light conditions, when the brightness information in the environment image is relatively sufficient and changes less, and the target model can mainly rely on position features to evaluate the recognition quality.

[0103] Specifically, the target model extracts position-related features such as the bounding box accuracy, contour integrity, and geometric shape regularity of the recognized object. The system evaluates the detection accuracy by calculating the fitting degree of the bounding box and the real object boundary. When the bounding box can closely fit the actual contour of the recognized object, it indicates that the position recognition accuracy is high. At the same time, the target model also analyzes the contour continuity and geometric consistency of the recognized object, and the object with complete and clear contour and expected geometric features will obtain a higher position confidence score.

[0104] In another possible implementation, the target model predicts confidence based solely on the brightness features of the identified object in the ambient image. This approach is primarily suitable for driving scenarios where position information may be unstable, such as in dynamic environments like high-speed driving, in urban road environments with significant occlusion, or in visually blurred conditions caused by rain or fog. In these driving scenarios, the position features of the identified object may be unreliable due to motion blur, partial occlusion, or image quality degradation. In this case, brightness features become an important basis for evaluating recognition quality.

[0105] Specifically, the target model calculates brightness-related parameters such as the brightness uniformity of the object area being identified, the contrast with the background, and the suitability of the lighting conditions. When the identified object has good brightness distribution and sufficient contrast, the target model can still accurately estimate the distance based on clear brightness information, even when the position characteristics are less than ideal. The system evaluates visual quality by analyzing the brightness gradient and texture characteristics within the object area. Objects with moderate brightness and clear texture receive higher brightness confidence scores.

[0106] In another possible implementation, the target model uses a combination of the position and brightness features of the identified object to predict confidence. This approach is suitable for complex driving environments where high recognition accuracy is required, such as nighttime driving, inclement weather conditions, or complex road environments in urban and rural areas. In these driving scenarios, a single feature may not provide a sufficiently reliable confidence assessment, and multiple feature fusion is needed to improve the accuracy and robustness of the assessment.

[0107] Specifically, the target model first extracts and quantifies the quality indicators of position features and brightness features separately, and then comprehensively analyzes these two types of features. The system can use a weighted distribution method to dynamically adjust the weight ratio of position features and brightness features in the confidence calculation based on the current environmental conditions. For example, in a well-lit environment, the system will increase the weight of position features, relying more on geometric shape and boundary information; while in low-light environments, the system will increase the weight of brightness features, focusing more on object visibility and contrast information. Through this approach, the target model can provide stable and reliable confidence predictions in a variety of complex environments.

[0108] It should be noted that the above-mentioned different implementations can be selected and configured based on actual application needs and system design requirements. In a high-performance autonomous driving system, multiple confidence prediction methods can be supported simultaneously, and the most appropriate prediction mode can be automatically switched based on real-time environmental monitoring results. In addition, different implementations can also be applied to different types of identified objects. For example, for vehicles, position features can be prioritized, while for pedestrians, brightness features can be prioritized, thereby achieving more refined confidence assessment and headlight control strategies.

[0109] On the basis of the above embodiment, the accurate evaluation of the vehicle perception capability and the determination of the headlight control parameters can also be achieved through confidence screening and maximum distance selection.

[0110] In operation S201, the confidence threshold can be determined in a variety of ways. In one embodiment, a fixed confidence threshold can be set based on the historical performance data and verification results of the target model. In another embodiment, the confidence threshold can be dynamically adjusted based on current environmental conditions. For example, a higher confidence threshold, such as 0.8, can be set in a well-lit daytime environment, while the confidence threshold can be appropriately lowered to 0.6 at night or in inclement weather conditions to adapt to changes in the recognition performance of the target model in different environments.

[0111] After determining the confidence threshold, the system traverses all initial distances and their corresponding confidence values ​​output by the target model, comparing them one by one. For each initial distance, the system checks whether its confidence exceeds the set confidence threshold. Initial distances that meet the criteria are marked as candidate distances and added to the candidate distance set. This effectively eliminates low-confidence distance data caused by poor image quality, object occlusion, or blurred recognition.

[0112] In operation S202, when the candidate distance set contains multiple distance values, the candidate distances may be sorted and compared, and the candidate distance with the largest value may be identified as the target distance. A numerical comparison algorithm may be used to traverse the candidate distance set, maintain a current maximum distance variable, and update the value of this variable by comparing each value one by one, ultimately determining the maximum distance value in the set.

[0113] It should be noted that when the set of alternative distances is empty, that is, the confidence levels of all initial distances are less than or equal to the confidence threshold, the system can use the preset default distance value as the target distance, or appropriately lower the confidence threshold and re-screen to ensure that the headlight control system can obtain the necessary distance parameters.

[0114] Figure 2 A schematic diagram of a scenario of a vehicle light control method provided in an embodiment of the present disclosure.

[0115] like Figure 2 As shown, a processing interface 201 of the vehicle perception system is displayed on an onboard display screen 200. To the left of the processing interface 201 is an environment image display area 202, which displays a road environment image currently captured by an onboard camera 203. The environment image includes vehicles 205A and 205B ahead, pedestrians 205C and 205D on the roadside, and a traffic sign 205E (the identified object).

[0116] To the right of processing interface 201 is distance data processing area 206, which displays the distance data entries corresponding to the identified objects. Each entry includes a distance value and a confidence level, such as "120m|0.9 (A)", "35m|0.5 (B)", "8m|0.8 (C)", "20m|0.7 (D)", and "50m|0.4 (E)".

[0117] Confidence threshold control bar 207, located below distance data processing area 206, displays the currently set confidence threshold of 0.6. Based on this threshold, distance data entries with a confidence level above the threshold are activated. Simultaneously, recognized objects with a confidence level above the threshold are assigned a qualified icon in the environment image display area.

[0118] On the right side of the processing interface 201 is a result output area 315 , in which a target distance display box 316 highlights the final determined target distance “12 m” with a highlighted border 317 . At the same time, a headlight control signal is generated based on the target distance.

[0119] By employing the disclosed embodiments and using confidence thresholds to filter multiple initial distances, low-quality recognition results can be effectively eliminated, ensuring sufficient reliability of the distance information used for headlight control. By selecting the maximum distance among the candidate distances as the target distance, the vehicle's maximum effective sensing capability in the current environment can be accurately reflected, ensuring that the headlight illumination range meets the vehicle's actual sensing needs, avoiding safety hazards caused by insufficient sensing capabilities and improving driving safety in complex road environments.

[0120] In actual vehicle driving, the vehicle's driving speed directly affects the perception distance and reaction time required for safe driving. If the lighting is adjusted only based on the environmental perception results, it will lead to the problem that the lighting range under different speed conditions does not match the actual safety requirements.

[0121] For example, when a vehicle is driving at high speed, even if the target model can detect objects at a certain distance under current lighting conditions, the actual required safe perception distance may be far greater than the current perception capability due to the increased braking distance and reaction time. In this case, if the headlight illumination range cannot be expanded accordingly, it will pose a safety hazard. Conversely, at low speeds, excessive illumination range not only wastes energy but also may cause glare to other road users.

[0122] To solve the above problem, based on the above embodiment, as an optional embodiment, operation S104 may further include the following operations:

[0123] Operation S301: obtaining a vehicle driving speed;

[0124] In operation S302 , a light control signal is output based on the vehicle driving speed and the target distance.

[0125] In operation S301 , the vehicle driving speed refers to the current real-time moving speed value of the vehicle, which is used to evaluate the minimum safety perception distance and headlight illumination range required by the vehicle at the current speed.

[0126] Vehicle speed can be acquired through the vehicle's onboard sensor systems, including but not limited to wheel speed sensors, GPS positioning systems, and inertial measurement units. These sensors can monitor the vehicle's motion in real time and provide accurate speed data. The frequency of vehicle speed acquisition can be adjusted based on system requirements. Typically, the frequency is the same or similar to that used for acquiring environmental images to ensure time synchronization between speed information and sensory information.

[0127] In operation S302 , the system inputs the vehicle driving speed and the environment image into the target model simultaneously, so that the target model can fully consider the current driving state of the vehicle when performing environment perception analysis.

[0128] Specifically, after receiving the environmental image and vehicle speed, the target model adjusts its internal perception processing based on this speed information. At high speeds, the target model enhances the recognition accuracy of distant targets and the tracking of fast-moving targets, while also improving the accuracy of longitudinal distance measurement to meet the safety requirements of longer perception distances in high-speed driving scenarios. At lower speeds, the target model focuses more on detailed recognition of close-range targets and coverage of the lateral perception range, accommodating the more comprehensive perception of the surrounding environment required at lower speeds.

[0129] Through this speed-guided perception optimization, the target model can output initial distance and confidence information that better meets the current driving state requirements, so that the target distance subsequently determined through confidence screening more accurately reflects the vehicle's actual perception capabilities at the current speed.

[0130] Furthermore, the system calculates the theoretical safety perception distance requirement based on the vehicle's driving speed and compares it with the target distance after speed optimization. When the target distance is still less than the safety perception distance requirement, the system will generate a lighting control signal to enhance the illumination range. When the target distance can meet or exceed the safety perception distance requirement, it can generate a lighting control signal to maintain the current illumination state.

[0131] By adopting the disclosed embodiments, the vehicle's driving speed is used as an input parameter for the target model and then integrated with the environmental image. This allows the target model to adaptively adjust its perception strategy based on the vehicle's dynamic driving state, resolving the mismatch between static image analysis and dynamic driving requirements in related perception methods. This improves the accuracy and safety of the headlight control system under varying driving speeds.

[0132] In actual vehicle driving, the vehicle's driving speed is not only affected by the driver's personal driving habits, but also needs to comply with road regulations and adapt to actual road conditions. If the headlights are controlled only based on the current actual driving speed, the headlight illumination range may not meet regulatory safety requirements in certain specific scenarios.

[0133] To address the above problem, based on the above embodiment, as an optional embodiment, operation S104 may further include the following operations:

[0134] Operation S401: obtaining speed limit information; the speed limit information includes at least one of road condition information of the vehicle's current driving road, regulatory information, and preset information input by the user;

[0135] Operation S402: determining a minimum driving speed based on speed limit information;

[0136] In operation S403 , a light control signal is output based on the minimum driving speed and the target distance.

[0137] In operation S401, speed limit information refers to various constraints and reference standards that affect the reasonable driving speed of the vehicle. In the embodiment of the present disclosure, it can be understood as comprehensive information used to determine the driving speed range that the vehicle should maintain under the current road environment, and is used to guide the vehicle's speed control and the corresponding adjustment of the headlight illumination range.

[0138] Optionally, the system can obtain speed limit information in a variety of ways, including electronic map data read by the on-board navigation system, road sign information recognized by the on-board camera, real-time traffic information received by the Internet of Vehicles system, and personal driving preference settings entered by the driver through the on-board human-computer interaction interface.

[0139] Road condition information includes real-time data such as the current road type, road grade, traffic flow, weather conditions, and construction restrictions. Regulatory information includes the legal maximum and minimum speed limits for the current road, temporary speed limits on special roads, and differentiated speed limits for different vehicle types. User-entered preset information includes driver-defined cruising speeds, personal driving preferences, safety margin settings, and other personalized parameters.

[0140] In operation S402, the minimum driving speed refers to the minimum driving speed value that the vehicle should maintain under the current environment after comprehensively considering the speed limit information. It is used to indicate the speed reference parameter for controlling the headlight illumination range to ensure that the headlight illumination capability can meet the safety requirements of the vehicle under various reasonable driving conditions.

[0141] Specifically, the system determines the minimum driving speed by analyzing the obtained speed limit information. When the speed limit information contains regulatory information, the system first extracts the statutory minimum speed limit of the current road as the basic reference value to ensure that the determined minimum driving speed is not lower than the minimum value required by law.

[0142] When the speed limit information includes road condition information, the system will modify the basic reference value based on actual traffic flow, road conditions and weather conditions. For example, in the case of traffic congestion, the expected minimum driving speed value may be appropriately lowered, and in the case of good road conditions, the minimum driving speed standard may be maintained or appropriately increased.

[0143] When speed limit information includes user-entered presets, the system balances the driver's personal preferences with regulatory requirements and road conditions, taking into account the driver's driving habits and comfort needs while meeting regulatory and safety requirements. The system comprehensively analyzes various speed limit information through weighted calculations or priority determinations to determine a minimum driving speed that meets regulatory requirements and adapts to actual road conditions.

[0144] In operation S403, the system calculates the theoretical braking distance and safe reaction distance for the vehicle at the minimum driving speed, and determines the minimum perception distance required by the vehicle. If the target distance is less than the minimum perception distance calculated based on the minimum driving speed, it indicates that the vehicle's perception capability under the current ambient lighting conditions cannot meet the minimum driving speed safety requirement. The system then generates a lighting control signal to increase the illumination range, instructing the lighting system to expand the illumination range to compensate for the insufficient perception capability.

[0145] When the target distance is greater than or equal to the minimum perception distance requirement, it indicates that the current perception capability can support safe driving at the lowest driving speed. The system will generate a lighting control signal to maintain or moderately adjust the illumination range to ensure that the illumination effect can meet safety requirements without causing excessive illumination.

[0146] In addition, the system will also consider the difference between the minimum driving speed and the vehicle's current actual driving speed. When the actual driving speed is significantly lower than the minimum driving speed, the system will prioritize the illumination range control according to the minimum driving speed standard, and prepare sufficient perception capabilities in advance for possible acceleration behavior of the vehicle.

[0147] By adopting the disclosed embodiments, the minimum driving speed is determined by comprehensively considering road regulations, actual road conditions, and the driver's personal preferences. This allows the headlight illumination range to be controlled not only based on the current actual perception conditions but also to meet the vehicle's reasonable driving needs under the current road conditions. By presetting the illumination range based on the minimum driving speed, the vehicle is guaranteed to have sufficient visual perception support when increasing driving speed or responding to emergencies, thereby improving driving safety and regulatory compliance.

[0148] During actual vehicle driving, the vehicle's motion state and driving trajectory will directly affect the vehicle's perception needs and focus of the surrounding environment. If the headlights are controlled only based on the distance information of all identified objects in the environmental image, the headlight illumination range will not match the vehicle's actual driving intention.

[0149] For example, when a vehicle is changing lanes, it primarily needs to focus on traffic conditions and safe distances to the side. If the headlight control system still adjusts its illumination range based on distant objects directly ahead, it will not be able to provide the necessary side lighting support for the lane change. Similarly, in different driving scenarios, such as high-speed straight-line driving and low-speed cornering, the vehicle's environmental perception requirements vary in terms of spatial scope and distance. A fixed perception processing approach cannot meet these dynamically changing driving needs.

[0150] To address the above problem, based on the above embodiment, as an optional embodiment, the above vehicle light control method may further include the following operations:

[0151] Operation S501: Acquire driving information; the driving information is determined based on the current motion trajectory and driving speed of the vehicle;

[0152] Operation S502: determining a target area based on driving information;

[0153] Operation S503: screening at least one candidate distance from the initial distances corresponding to the identified objects located in the target area;

[0154] In operation S504 , a target distance is determined according to each candidate distance and the corresponding confidence level.

[0155] In operation S501, driving information refers to data reflecting the current motion state and driving intention of the vehicle, which can be understood in the embodiment of the present disclosure as a set of parameters calculated based on the vehicle motion trajectory and driving speed to describe the dynamic behavior characteristics of the vehicle.

[0156] For example, the system can monitor the changes in the vehicle's motion trajectory in real time through on-board sensors, including the vehicle's driving direction, steering angle, lateral displacement, longitudinal acceleration and other motion parameters, and combine the vehicle's current driving speed information to comprehensively analyze the vehicle's motion state and possible driving intentions.

[0157] Specifically, when the vehicle is driving in a straight line at a high speed, the driving information will indicate that the vehicle is in a high-speed, straight-line driving state, and the vehicle should focus on the long-distance traffic conditions directly ahead. When the vehicle is turning or moving laterally, the driving information will indicate that the vehicle is changing lanes or turning, and the vehicle should focus on the traffic environment in the turning direction or the area to the side.

[0158] In operation S502, the target area refers to the spatial range that needs to be focused on and perceived based on the vehicle driving information. In the embodiment of the present disclosure, it can be understood as the environmental perception area that is most relevant to the vehicle's current driving state and intention, and is used to instruct the target model to perform targeted target recognition and distance analysis.

[0159] Specifically, when driving information indicates that the vehicle is in a high-speed straight-line driving state, the system will set the long-distance area directly in front as the target area, which covers a larger longitudinal distance range in front of the vehicle to meet the safety requirements of long-distance perception during high-speed driving.

[0160] When driving information indicates that the vehicle is in a low-speed driving state, the system will set a relatively small area directly in front as the target area, focusing on close-range traffic conditions and obstacles.

[0161] When driving information indicates that the vehicle is changing lanes, the system will set the side area in the direction of lane change as the target area to ensure that the vehicle can obtain the necessary side traffic information during the lane change process.

[0162] When driving information indicates that the vehicle is turning, the system will set the curve area in front of the turning direction as the target area, including the spatial range inside the curve and at the curve exit, to adapt to the vision requirements during turning driving.

[0163] By dynamically adjusting the position and size of the target area, the system can achieve regionalized perception processing that matches the vehicle's driving status.

[0164] In operation S503 , the system performs spatial position analysis on all recognized objects identified by the target model to determine whether each recognized object is located within the range of the target area.

[0165] For identified objects within the target area, the system extracts the corresponding initial distance value and adds it to the candidate distance set as a candidate distance. For identified objects outside the target area, the system excludes their corresponding initial distance from the candidate distance set to prevent distance information unrelated to the current driving intention from interfering with headlight control.

[0166] Through the above screening process, the system can ensure that the distance information contained in the candidate distance set is highly relevant to the vehicle's current driving needs and focus, thereby improving the pertinence of the distance information.

[0167] It should be noted that when there is no identified object in the target area, the system can expand the scope of the target area for re-screening, or use a preset default candidate distance value to ensure that the headlight control system can obtain the necessary distance parameters.

[0168] In operation S504, the system may use the same confidence screening method as the above embodiment, set a confidence threshold and screen out distance values ​​with confidence greater than the threshold from the candidate distances, and then select the maximum distance from the screened distances as the target distance.

[0169] Furthermore, the system can employ differentiated target distance determination strategies based on the characteristics of different driving states. For example, during high-speed, straight-line driving, it prioritizes the maximum candidate distance to ensure adequate long-range perception. During lane changes, it prioritizes moderately sized candidate distances with high confidence levels to balance perception accuracy and reaction time requirements. By combining driving state characteristics with confidence information, the system can determine a target distance that both meets current driving needs and provides sufficient reliability.

[0170] By employing the disclosed embodiments, driving information is determined based on the vehicle's trajectory and driving speed, and the target area most relevant to the current driving state is dynamically determined based on this driving information. This allows target recognition and distance analysis to focus on the spatial range of the vehicle's actual interest. By selecting candidate distances from identified objects within the target area and determining the target distance, the adjustment of the headlight illumination range accurately responds to the vehicle's dynamic driving needs, improving the adaptability and accuracy of the headlight control system in various driving scenarios.

[0171] During actual vehicle driving, the headlight control methods in related technologies often adjust illumination based on preset fixed strategies or simple ambient light sensing. This method cannot accurately assess the degree of match between the vehicle's actual perception needs under current driving conditions and the actual perception capabilities of the target model, resulting in the problem that the headlight illumination range may deviate from actual needs.

[0172] For example, in a high-speed driving scenario at night, the vehicle requires a longer perception distance to ensure sufficient braking reaction time due to the high driving speed. At the same time, due to the low ambient brightness, the perception ability of the target model will decrease accordingly. At this time, if the headlight control system cannot quantify the gap between demand and ability, insufficient or excessive illumination may occur.

[0173] Conversely, during daytime, low-speed driving, even when ambient brightness is sufficient, if the system still controls illumination according to nighttime standards, it will result in unnecessary energy waste and potential interference to other road users. Related technologies lack a quantitative analysis of the vehicle's actual perception needs and the target model's true perception capabilities, making it impossible to achieve precise, demand-driven headlight control.

[0174] To address the above problem, based on the above embodiment, as an optional embodiment, the above vehicle light control method may further include the following operations:

[0175] Operation S601: obtaining a vehicle driving speed and an ambient brightness;

[0176] Operation S602: determining a first perception range based on the vehicle's driving speed and the ambient brightness change, where the first perception range represents an area that the vehicle needs to recognize under the ambient brightness;

[0177] Operation S603: determining a second perception range based on the multiple initial distances, where the second perception range represents an area where the target model can be identified under ambient brightness;

[0178] Operation S604: output a light control signal based on the first sensing range and the second sensing range.

[0179] In operation S601, the ambient brightness directly affects the image processing quality and target recognition accuracy of the target model. In a low-brightness environment, image noise increases and contrast decreases, which significantly weakens the perception ability and reliable perception distance of the target model.

[0180] For example, the system can obtain vehicle speed information in real time through onboard sensors. Ambient brightness can be acquired through onboard illumination sensors, camera automatic exposure parameters, or statistical analysis of image brightness. The system must ensure that the acquisition frequency of vehicle speed and ambient brightness parameters is synchronized with the update frequency of the ambient image to ensure temporal consistency of the analysis.

[0181] In operation S602, the first perception range refers to the minimum perception area coverage that the vehicle must have to ensure safe driving, calculated based on the vehicle's current driving speed and ambient brightness conditions. In the embodiment of the present disclosure, it can be understood as a theoretical perception requirement area determined after comprehensively considering the vehicle's dynamic characteristics, ambient visual conditions and safety margin requirements.

[0182] Exemplarily, the first perception range may be a rectangular area starting from the current position of the vehicle, extending forward and extending left and right. The first perception range can cover the longitudinal distance required for safe braking of the vehicle at the current speed and the lateral range required for possible lane changing operations.

[0183] In addition, the first perception range can also be a fan-shaped area determined according to the road curvature characteristics. When the vehicle is driving on a curve, the perception range will be adjusted accordingly along the direction of the road curve to adapt to the field of vision requirements when driving on a curve.

[0184] In a feasible implementation, the calculation of the first perception range may further include the following operations:

[0185] S701, determining a first longitudinal distance required for the vehicle to be recognized under ambient brightness based on the vehicle driving speed;

[0186] S702, determining a first lateral distance that the vehicle needs to be identified under ambient brightness based on the road the vehicle is on;

[0187] S703: Determine a first perception range based on the first longitudinal distance and the first lateral distance.

[0188] In operation S701 , the first longitudinal distance refers to a minimum perception distance that the vehicle must have in the longitudinal direction to ensure safe driving, which is calculated based on the current driving speed of the vehicle and ambient brightness conditions.

[0189] In one possible implementation, the system calculates a theoretical braking distance based on the vehicle's driving speed and then modifies it based on ambient brightness. When ambient brightness is low, the system needs to add a safety margin to the theoretical braking distance due to increased delays and uncertainty in visual recognition. The lower the ambient brightness, the greater the required safety margin.

[0190] In another feasible implementation, the first longitudinal distance can be broken down into a reaction distance and a braking distance. The reaction distance dynamically adjusts the reaction time parameter based on ambient brightness, with lower ambient brightness resulting in a longer reaction time. The braking distance adjusts the road friction coefficient based on ambient brightness, using a more conservative friction coefficient value in low-light environments. The first longitudinal distance is equal to the sum of the reaction distance and the corrected braking distance.

[0191] Alternatively, a two-dimensional lookup table can be created for vehicle speed and ambient brightness, dividing the speed range and ambient brightness into multiple intervals and levels, respectively, and presetting corresponding first longitudinal distance reference values ​​for different combinations. When actual conditions do not completely correspond to the discrete values ​​in the lookup table, bilinear interpolation is used to calculate the first longitudinal distance.

[0192] In operation S702 , the first lateral distance refers to a required perception range of the vehicle in the lateral direction determined based on the current road characteristics and ambient brightness conditions of the vehicle.

[0193] Specifically, the system first obtains road information, including the number of lanes, lane width, and lane change availability. For single-lane roads, the first lateral distance primarily considers the safety margin at the lane boundary. For multi-lane roads where lane changes are permitted, the first lateral distance must account for adjacent lanes, including their widths and necessary safety gaps. The system then adjusts the calculated results based on ambient brightness, requiring an additional lateral safety margin in low-light environments.

[0194] In operation S703, a vehicle coordinate system may be established with the current position of the vehicle as the coordinate origin, and the first perception range may be constructed as a rectangular area. Whether the target object's coordinate position falls within the rectangular area is determined to determine whether the target object falls within the perception requirement range.

[0195] For curved roads or complex paths, the system can use a sector-shaped area to more accurately describe the sensing range. The sector's radius is equal to the first longitudinal distance, and the sector's angular range is calculated based on the ratio of the first lateral distance to the longitudinal distance. When the vehicle is driving on a curve, the sector's central axis will shift in angle based on the road curvature.

[0196] The system can also divide the primary sensing range into multiple sub-areas, each with different priorities and perception accuracy requirements. The core area directly in front of the vehicle requires the highest perception accuracy, while the extended area covering the rest of the area can appropriately reduce accuracy requirements. Through area identification and priority management, differentiated processing of targets within different spatial ranges is achieved.

[0197] It should be noted that the predictability of ambient brightness changes also needs to be considered when determining the first perception range. Because ambient brightness can change in a short period of time, such as when entering a tunnel, passing under the shadow of a bridge, or experiencing sudden changes in weather conditions, the calculation of the first perception range should not only be based on the current ambient brightness state, but also consider the impact of possible brightness change trends on perception needs. Furthermore, ambient brightness changes in different time periods have different characteristics. The system can predict the changing pattern of ambient brightness based on time information and historical data, and incorporate this predictive information into the calculation of the first perception range to improve the accuracy and adaptability of the perception range setting.

[0198] In operation S603, the second perception range refers to the spatial area range in which the target model can actually be reliably identified and measured under the current environmental brightness conditions, which is obtained based on the statistical analysis of multiple initial distances output by the target model. In the embodiment of the present disclosure, it can be understood as an objective evaluation area that reflects the true perception ability state of the target model, and is used to quantify the effective perception boundary and recognition distribution of the target model under the current environmental conditions.

[0199] For example, the second perception range can be a fan-shaped area derived from the target model's past recognition results, representing the maximum spatial range within which the target model can stably and reliably identify the target under the current ambient brightness. The second perception range can also be a multi-layered area divided according to different confidence levels, such as a high-reliability area with a confidence level greater than 0.9, a medium-reliability area with a confidence level greater than 0.7, and a basic-reliability area with a confidence level greater than 0.5. This layered description more comprehensively reflects the distribution characteristics of the target model's perception capabilities.

[0200] In a feasible implementation, a statistical analysis method can be used to determine the second perception range based on multiple initial distances. First, all the initial distances output by the target model within a certain time window and their corresponding confidence levels are obtained. By setting a confidence threshold, the system retains the initial distances with a confidence level greater than the threshold as reliable recognition results, performs a distribution analysis on the distance data of the reliable recognition results, and calculates statistical features such as the maximum value, average value, median, and standard deviation of the distance data. The maximum value or percentile value of the reliable recognition distance is used as the boundary distance of the second perception range, and the angular coverage of the perception range is determined based on the distribution of reliably recognized targets at different directions and angles.

[0201] In another feasible implementation, a dynamic update method can be used to determine the second perception range, which can reflect the changing trends of the target model's perception capabilities in real time. The system maintains a sliding time window, continuously collecting and updating the recognition result data of the target model. When new initial distance data is input, the system adds the new data to the analysis set and removes expired data outside the time window to ensure the timeliness of the analysis data. By recalculating the updated data set, the latest second perception range parameters are obtained, and the new perception range is compared with the previous calculation results to analyze the changing trends and stability of the perception capabilities.

[0202] When perception changes significantly, the system adjusts the size and shape of the second perception range accordingly. When perception remains stable, the system maintains the current perception range setting. This allows the second perception range to respond promptly to changes in environmental conditions and fluctuations in target model performance.

[0203] In operation S604, the specific requirements for vehicle headlight control are determined by comparing and analyzing the differences between the first and second perception ranges. Based on this difference analysis, corresponding light control signals are output. First, the spatial difference between the first and second perception ranges is calculated, including the longitudinal distance difference and the lateral angle difference. The longitudinal distance difference reflects the target model's perceived gap in straight-line distance, while the lateral angle difference reflects the target model's perceived limitations in field of view. Based on the analysis results, the first and second control signals are generated.

[0204] The first control signal is used to control the brightness of the headlights. This signal is primarily determined based on the difference in longitudinal distance between the first and second sensing ranges. When the longitudinal distance of the second sensing range is significantly smaller than that of the first sensing range, this indicates that the target model's recognition distance is insufficient under the current ambient brightness. Increasing the headlight brightness to improve the brightness level of the illuminated area will improve the image input quality and recognition performance of the target model. The first control signal includes specific brightness adjustment parameters, such as a brightness increase percentage, a target illuminance value, or a modulation parameter.

[0205] Similarly, the second control signal is used to control the illumination angle of the headlights. This signal is primarily determined based on the difference in lateral coverage between the first and second sensing ranges. When the lateral angle coverage of the second sensing range is less than the lateral angle requirement of the first sensing range, it indicates that the field of view of the target model under the current illumination conditions is insufficient, and it is necessary to expand the lateral coverage of the illuminated area by adjusting the illumination angle of the headlights. The second control signal includes specific angle adjustment parameters, such as the horizontal illumination angle increment, the vertical illumination angle adjustment amount, or the illumination direction offset angle.

[0206] It should be noted that the generation of the first and second control signals also needs to consider the physical limitations of the lighting system and regulatory requirements. Lamp brightness must be adjusted within the maximum permitted by law to avoid excessive glare for other road users. Adjustment of the illumination angle must also consider the mechanical adjustment range and response speed of the lamp to ensure feasibility and real-time performance of angle adjustment.

[0207] By adopting the embodiments of the present disclosure, the system can quantify in real time the vehicle's perception needs under current driving conditions and the actual perception capabilities of the target model, and accurately determine the adjustment needs of the headlight illumination by comparing the differences between the two perception ranges, so that the headlight illumination range can accurately match the actual safety needs of the vehicle, thereby avoiding safety hazards caused by insufficient perception capabilities, preventing energy waste and interference to other road users caused by excessive illumination, and improving the accuracy and adaptability of the headlight control system under different driving conditions.

[0208] On the basis of the above-mentioned embodiments, as an optional embodiment, the target perception range is determined based on the difference between the first perception range and the second perception range, and the target perception range represents the perception capability of the target model under the driving speed of the vehicle and the ambient brightness.

[0209] The target perception range refers to an evaluation index for quantifying the perception capability of the target model under specific driving speed of the vehicle and ambient brightness conditions, which is calculated based on the difference between the first perception range and the second perception range. In the embodiments of the present disclosure, it can be understood as a differentiated measurement parameter reflecting the matching degree between the actual perception capability of the target model and the theoretical perception demand, which is used to accurately evaluate the performance state of the automatic driving system, guide the control decision of the vehicle light, optimize the allocation of computing resources, etc.

[0210] In specific applications, the target perception range can be used as a control parameter of the intelligent lighting system of the vehicle. By monitoring the changes of the target perception range in real time, the brightness and irradiation angle of the vehicle light can be dynamically adjusted to match the perception capability and the perception demand, actively improve the illumination environment of the road, and enable the driving vehicle to obtain clear visual perception effect in any scene, thereby improving the visual detection accuracy.

[0211] The target perception range can also be applied to the health monitoring and fault warning of the perception system. When the target perception range abnormally fluctuates or continuously deviates from the normal range, the system can timely discover the trend of the decline of the perception performance and trigger the corresponding maintenance or calibration program.

[0212] In a feasible implementation, the first perception range and the second perception range can be represented as multi-dimensional vectors, each dimension corresponding to a different spatial direction or distance parameter, such as longitudinal distance, left lateral distance, right lateral distance, upper vertical angle, and lower vertical angle, etc. The second perception range vector is subtracted from the first perception range vector to obtain a difference vector, and each component of the difference vector represents the perception capability surplus or gap of the target model in the corresponding spatial direction. The length of the difference vector is further calculated as the target perception range, and the size of the length reflects the overall degree of the perception capability deviation. The sign of the length is determined by the weighted average of the components of the difference vector, and the positive and negative values represent the overall perception capability. Through the above vector processing method, the target perception range can retain detailed information of the perception capability difference in different spatial directions.

[0213] In another feasible implementation, a weighted fusion method can be used to determine the target perception range, so that each perception dimension can be treated differently according to the characteristics of different driving scenarios. The system sets the weight coefficients of different perception dimensions according to the current vehicle driving speed, road type and traffic conditions, giving a higher weight to the longitudinal perception distance in highway scenarios, a higher weight to the lateral field of view coverage in complex urban intersection scenarios, and adding additional safety weights to all perception dimensions in nighttime or bad weather scenarios. The weighted difference between the first perception range and the second perception range in each dimension is calculated, and then the weighted difference of each dimension is combined to obtain the target perception range. Ensure that the final target perception range can accurately reflect the perception capability evaluation results and control demand priorities in the current driving scenario.

[0214] By adopting the embodiments of the present disclosure, the system can accurately quantify the perception capability status of the target model under specific vehicle driving speed and ambient brightness conditions. It can not only identify situations where perception capability is insufficient so as to take compensatory measures in a timely manner, but also discover situations where perception capability is excessive so as to optimize resource allocation.

[0215] Based on the above embodiments, the present disclosure also provides a vehicle light control system. Figure 3 , Figure 3 The following schematically shows a block diagram of the architecture of a vehicle light control system according to an embodiment of the present disclosure.

[0216] like Figure 3 As shown, the headlight control system adopts a modular layered architecture design, and realizes the process from environmental data collection to headlight control execution through the collaborative work of the acquisition module 310, the automatic driving module 320, the lighting control module 330 and the headlight module 340.

[0217] The acquisition module 310 is the data input terminal of the system and is responsible for obtaining basic environmental information during the vehicle's driving process. This module collects environmental images through the vehicle's onboard camera and also obtains key parameters such as vehicle driving speed and ambient brightness.

[0218] The autonomous driving module 320 is the system's processing module, integrating three units: an environmental assessment model 321, an optimal constraint unit 322, and an environmental demand model 323. The environmental assessment model 321 receives the environmental image from the acquisition module 310, identifies and measures the distance of the identified objects in the image using the target model, outputs multiple initial distances and their corresponding confidence levels, and determines the target distance based on the confidence level, ultimately evaluating the actual perception capability of the target model. The optimal constraint unit 322 is responsible for comprehensively analyzing information such as driver input, regulatory requirements, and road conditions to determine the constraints for system operation, including parameters such as the minimum operating speed and vehicle motion range. The environmental demand model 323 is a decision-making unit that receives the actual perception capability assessment results provided by the environmental assessment model 321 and the system constraints determined by the optimal constraint unit 322. By comparing and analyzing the differences between the actual perception capability and the theoretical perception requirements, it determines the adjustment requirements for the headlight illumination range and generates the corresponding lighting control signals.

[0219] The lighting control unit 331 in the lighting control module 330 receives the lighting control signal from the environmental demand model 323, and calculates specific vehicle light control parameters based on the differences in perception capabilities and illumination requirements, including control instructions such as illumination distance, illumination angle, and light intensity, to ensure that the generated control parameters can meet both perception requirements and regulatory requirements.

[0220] The headlight module 340 is the execution terminal of the system, which receives the control instructions output by the light control unit 331 and realizes precise control of the illumination range, brightness and angle of the headlight through the headlight driving circuit and adjustment mechanism.

[0221] exist Figure 3 Based on Figure 4 , Figure 4 The following schematically shows a signaling interaction diagram under a vehicle light control system provided according to an embodiment of the present disclosure.

[0222] Operation S701: the acquisition module 310 acquires an environment image;

[0223] Operation S702: The acquisition module 310 sends an environmental image to the environmental assessment model 321;

[0224] Operation S703: The environment assessment model 321 analyzes and processes the environment image to determine a plurality of initial distances;

[0225] In operation S704, the environment assessment model 321 determines a target distance according to the confidence level of each initial distance.

[0226] Operation S705: The environment assessment model 321 sends an environment description signal to the environment requirement model 323. The environment description signal includes a target distance, which is used to characterize the current perception capability of the target model.

[0227] In operation S706, the optimal constraint unit 322 obtains system description information and regulatory requirement information. The system description information includes parameters such as the target speed input by the driver and the current driving speed of the vehicle. The regulatory requirement information includes road condition information such as the minimum operating speed, number of lanes, travel direction, and lane change conditions.

[0228] In operation S707 , the optimal constraint unit 322 determines the constraint conditions for system operation based on the system description information and the regulatory requirement information. The constraint conditions include the minimum driving speed and the vehicle movement range.

[0229] In operation S708 , the optimal constraint unit 322 sends an environmental demand signal to the environmental demand model 323 . The environmental demand signal includes information about the minimum driving speed and vehicle motion range, which is used to represent the system's expected performance requirements for perception capabilities.

[0230] Operation S709: The environment requirement model 323 generates a lighting requirement signal according to the environment description signal and the environment requirement signal. The lighting requirement signal includes control parameters such as the illumination area range, illumination distance, and brightness requirement.

[0231] Operation S710: The environmental demand model 323 sends a lighting demand signal to the lighting control unit 331;

[0232] Operation S711: The light control unit 331 generates a vehicle light control instruction according to the light demand signal. The vehicle light control instruction includes a light brightness adjustment parameter, an illumination angle adjustment parameter, and an illumination range control parameter.

[0233] Operation S712: The light control unit 331 sends a light control instruction to the light module 340;

[0234] In operation S713 , the vehicle light module 340 executes the vehicle light control instruction through the vehicle light driving circuit and the adjustment mechanism;

[0235] It should be noted that the specific meaning of the above-mentioned signaling interaction process can be found in the implementation principles of the above-mentioned embodiment, and will not be elaborated on here. In addition, the above-mentioned signaling interaction process is only an exemplary interaction process of the above-mentioned headlight control system. In specific scenarios, the headlight control system can also be adjusted and optimized accordingly according to actual application requirements.

[0236] The embodiment of the present disclosure also discloses a vehicle light control device.

[0237] Figure 5 The figure schematically shows a structural block diagram of a vehicle light control device according to an embodiment of the present disclosure.

[0238] like Figure 5 As shown, the vehicle light control device includes:

[0239] An acquisition module 501 is used to acquire an environment image;

[0240] A first determination module 502 is configured to obtain a plurality of initial distances determined by the target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image;

[0241] A second determination module 503 is configured to determine a target distance based on a plurality of initial distances and a confidence level of each initial distance;

[0242] The output module 504 outputs a light control signal according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

[0243] Based on the above embodiment, as an optional embodiment, the second determination module 503 is further used to determine at least one alternative distance from multiple initial distances, where the alternative distance is an initial distance with a confidence level greater than a confidence level threshold; and the alternative distance with the largest distance among the at least one alternative distance is used as the target distance.

[0244] Based on the above embodiment, as an optional embodiment, the obtaining module 501 is further used to obtain the vehicle driving speed;

[0245] The output module 504 is further configured to output a light control signal based on the vehicle driving speed and the target distance.

[0246] Based on the above embodiment, as an optional embodiment, the obtaining module 501 is further configured to obtain speed limit information; the speed limit information includes at least one of road condition information of the vehicle's current driving road, regulatory information, and preset information input by the user;

[0247] The output module 504 is further configured to determine a minimum driving speed based on the speed limit information; and output a light control signal based on the minimum driving speed and the target distance.

[0248] Based on the above embodiment, as an optional embodiment, the obtaining module 501 is further configured to obtain driving information; the driving information is determined based on the current motion trajectory and driving speed of the vehicle;

[0249] The second determination module 503 is configured to determine a target area based on the driving information; select at least one candidate distance from the initial distances corresponding to the identified objects within the target area; and determine a target distance based on each candidate distance and the corresponding confidence level.

[0250] Based on the above embodiment, as an optional embodiment, the obtaining module 501 is further used to obtain the vehicle driving speed and the ambient brightness;

[0251] The second determining module 503 is further configured to determine a first perception range according to the driving speed of the vehicle and the ambient brightness, the first perception range representing a region that needs to be recognized by the vehicle under the ambient brightness; and determine a second perception range based on the plurality of initial distances, the second perception range representing a region that can be recognized by the target model under the ambient brightness.

[0252] The output module 504 is further configured to output a light control signal based on the first perception range and the second perception range.

[0253] In addition to the above embodiments, as an optional embodiment, the second determining module 503 is further configured to determine a target perception range based on a difference between the first perception range and the second perception range, the target perception range representing a perception capability of the target model under the driving speed of the vehicle and the ambient brightness.

[0254] In addition to the above embodiments, as an optional embodiment, the second determining module 503 is further configured to determine a first longitudinal distance that needs to be recognized by the vehicle under the ambient brightness according to the driving speed of the vehicle; determine a first lateral distance that needs to be recognized by the vehicle under the ambient brightness according to a road on which the vehicle is located; and determine the first perception range based on the first longitudinal distance and the first lateral distance.

[0255] Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure, or at least part of the functions of any one or more of the modules, sub-modules, units, sub-units, can be implemented in one module. Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be split into multiple modules for implementation. Any one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on board, a system on package, an application specific integrated circuit (ASIC), or any other reasonable hardware or firmware that can be integrated or packaged with a circuit, or implemented in any one of software, hardware, and firmware or in a proper combination of any one or more of them. Alternatively, one or more of the modules, sub-modules, units, sub-units according to the embodiments of the present disclosure can be at least partially implemented as computer program modules that can perform corresponding functions when the computer program modules are run.

[0256] For example, any multiple of the obtaining module 501, the first determining module 502, the second determining module 503, and the output module 504 can be combined into a single module / unit / sub-unit, or any one of these modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functionality of one or more of these modules / units / sub-units can be combined with at least part of the functionality of other modules / units / sub-units and implemented in a single module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the obtaining module 501, the first determining module 502, the second determining module 503, and the output module 504 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented in hardware or firmware by any other reasonable means of integrating or packaging circuits, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of these. Alternatively, at least one of the obtaining module 501 , the first determining module 502 , the second determining module 503 and the outputting module 504 may be at least partially implemented as a computer program module, which may perform corresponding functions when executed.

[0257] It should be noted that the data processing system part in the embodiments of the present disclosure corresponds to the data processing method part in the embodiments of the present disclosure. The description of the data processing system part specifically refers to the data processing method part and will not be repeated here.

[0258] The present disclosure also discloses a vehicle.

[0259] Figure 6 The figure schematically shows a structural block diagram of a vehicle according to an embodiment of the present disclosure.

[0260] like Figure 6 As shown, the vehicle 600 includes:

[0261] Vehicle body 601;

[0262] A vehicle light 602 is provided on the vehicle body 601;

[0263] The acquisition device 603 is provided on the vehicle body 601 and is used to acquire an image of the environment;

[0264] A first controller 604, disposed in the vehicle body 601, is configured to obtain a plurality of initial distances determined by the target model based on the environment image, the initial distances representing the corresponding distances of the identified objects in the environment image; determine a target distance based on the plurality of initial distances and the confidence level of each initial distance; and output a light control signal based on the target distance;

[0265] The second controller 605 is arranged on the vehicle body 601, configured to receive a light control signal, and control the illumination range of the vehicle lamp 602 according to the light control signal.

[0266] In the embodiments of the present disclosure, the collection device 603 can include but is not limited to a vehicle-mounted camera, an image sensor, a multi-view camera, and the like, which can collect image information of the road environment in front of and around the vehicle in real time. The first controller 604 can be a dedicated image processing controller or a processing unit integrated in a vehicle-mounted computing platform, which has sufficient computing power to run a target model and perform environment perception analysis. The second controller 605 can be a vehicle lamp control dedicated controller or a component of a vehicle body electronic control unit, which is responsible for receiving control instructions and driving the vehicle lamp actuator to adjust the illumination parameters.

[0267] The vehicle lamp 602 can include headlamps, fog lamps, turn signals, and the like, which have functions such as brightness adjustment, illumination angle adjustment, and illumination distance adjustment, and can accurately control the illumination range according to the received light control signal.

[0268] By using the vehicle provided by the embodiments of the present disclosure, intelligent vehicle lamp control based on environment perception and confidence analysis can be realized, so that the illumination range of the vehicle lamp can be dynamically adjusted according to the position and distance of the identified object in the actual driving environment, and the safety and adaptability of the vehicle in a complex driving environment are improved.

[0269] It should be noted that the first controller 604 in the vehicle 600 can perform the operations in the above vehicle lamp control method embodiments, and the specific technical implementation manner and technical effects can refer to the related descriptions of the above method embodiments, which will not be described here.

[0270] The embodiments of the present disclosure also include a computer program product, which includes a computer program containing program codes for executing the method provided by the embodiments of the present disclosure, and when the computer program product is running on an electronic device, the program codes are used to make the electronic device implement the method provided by the embodiments of the present disclosure.

[0271] When the computer program is executed by the processor, the above functions defined in the system / apparatus of the embodiments of the present disclosure are executed. According to the embodiments of the present disclosure, the system, apparatus, module, unit, etc. described above can be implemented by computer program modules.

[0272] In one embodiment, the computer program may be stored on a tangible storage medium, such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal over a network medium, downloaded and installed via a communication component, and / or installed from a removable medium. The program code contained in the computer program may be transmitted using any suitable network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0273] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the boxes may occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession may actually be executed substantially in parallel, or they may sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, as well as the combination of boxes in the block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or may be implemented using a combination of dedicated hardware and computer instructions. It will be understood by those skilled in the art that the features described in the various embodiments and / or claims of the present disclosure may be combined and / or coupled in various ways, even if such combinations or couplings are not explicitly described in the present disclosure. In particular, without departing from the spirit and teachings of the present disclosure, the features described in the various embodiments and / or claims of the present disclosure may be combined and / or coupled in various ways. All such combinations and / or couplings fall within the scope of the present disclosure.

[0274] The embodiments of the present disclosure are described above. However, the above embodiments are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although each embodiment is described above separately, it does not mean that the measures in each embodiment cannot be used in combination to advantage. The scope of the present disclosure is defined by the appended claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art may make various substitutions and modifications, and the above substitutions and modifications should all fall within the scope of the present disclosure.

Claims

1. A vehicle light control method, comprising: Get the environment image; Obtaining a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image; determining a target distance according to the multiple initial distances and a confidence level of each initial distance; A light control signal is output according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

2. The method according to claim 1, wherein the confidence level of the initial distance is obtained by predicting the target model based on the position of the identified object in the environment image and / or the brightness of the identified object; The determining the target distance according to the multiple initial distances and the confidence level of each initial distance includes: determining at least one candidate distance from the plurality of initial distances, the candidate distance being an initial distance having a confidence greater than a confidence threshold; The candidate distance with the largest distance among the at least one candidate distance is used as the target distance.

3. The method according to claim 1, wherein outputting a light control signal according to the target distance comprises: Get the vehicle driving speed; A light control signal is output based on the vehicle driving speed and the target distance.

4. The method according to claim 1, wherein outputting a light control signal according to the target distance comprises: Get speed limit information; The speed limit information includes at least one of road condition information of the vehicle's current driving road, regulatory information, and preset information input by a user; determining a minimum driving speed based on the speed limit information; A light control signal is output based on the minimum driving speed and the target distance.

5. The method according to claim 1, comprising: Get driving information; The driving information is determined based on the current motion trajectory of the vehicle and the driving speed of the vehicle; determining a target area based on the driving information; Filter at least one candidate distance from the initial distances corresponding to the identified objects located in the target area; The target distance is determined according to each candidate distance and the corresponding confidence level.

6. The method according to claim 1, comprising: Obtain vehicle driving speed and ambient brightness; determining a first perception range according to the vehicle driving speed and the ambient brightness change, where the first perception range represents an area that the vehicle needs to recognize under the ambient brightness; determining a second perception range based on the multiple initial distances, where the second perception range represents an area where the target model is recognizable under the ambient brightness; A light control signal is output based on the first sensing range and the second sensing range.

7. The method according to claim 6, comprising: A target perception range is determined based on a difference between the first perception range and the second perception range, where the target perception range represents a perception capability of the target model under the vehicle driving speed and the ambient brightness.

8. The method according to claim 6, wherein determining the first perception range according to the vehicle driving speed and the ambient brightness change comprises: determining a first longitudinal distance required to be identified for the vehicle under the ambient brightness according to the vehicle driving speed; Determining a first lateral distance of the vehicle that needs to be identified under the ambient brightness according to the road on which the vehicle is located; A first perception range is determined based on the first longitudinal distance and the first lateral distance.

9. A vehicle light control device, comprising: Acquisition module, used to obtain environmental images; A first determination module is configured to obtain a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image; a second determining module, configured to determine a target distance based on the multiple initial distances and a confidence level of each initial distance; The output module outputs a light control signal according to the target distance, and the light control signal is used to control the illumination range of the vehicle light.

10. A vehicle comprising: Vehicle body; A headlight, arranged on the vehicle body; A collection device, provided on the vehicle body, for acquiring an environmental image; a first controller, disposed in the vehicle body, configured to obtain a plurality of initial distances determined by a target model according to the environment image, wherein the initial distances represent corresponding distances of the identified objects in the environment image; determining a target distance according to the multiple initial distances and a confidence level of each of the initial distances; and outputting a light control signal according to the target distance; The second controller is provided in the vehicle body and is used for receiving the light control signal and controlling the illumination range of the vehicle light according to the light control signal.

Images