Vehicle Ambient Light Control Method and System Based on Environmental Perception

Abstract

The present invention relates to the field of intelligent control technology, and discloses a vehicle atmosphere lamp control method and system based on environmental perception. The method includes: collecting in-vehicle image data and vehicle-mounted sensor data and inputting them into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map; performing situation recognition and analysis on the fused feature map to obtain a driving situation feature vector and a situation correlation matrix; performing decoupling processing on the vehicle atmosphere lamp parameters according to the driving situation feature vector and the situation correlation matrix to obtain decoupled lighting parameters; performing situation-adaptive lighting regulation and dynamic situation weighted loss calculation on the decoupled lighting parameters and the driving situation feature vector to obtain target lighting control parameters. The present invention solves the problem of unbalanced lighting requirements in different driving scenarios and makes the adjustment of vehicle atmosphere lamps more accurate.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control technology, and particularly to a vehicle-mounted atmosphere lamp control method and system based on environmental perception. Background Art

[0002] As an important part of enhancing the user's driving experience, vehicle-mounted atmosphere lamps have been widely used in the field of intelligent vehicles in recent years. With the rapid development of intelligent connected vehicle technology, users' demands for personalization and intelligence of the in-vehicle environment are increasing day by day. Vehicle-mounted atmosphere lamps are no longer just simple decorative lighting, but also undertake multiple functions such as adjusting the driving atmosphere, prompting the driving state, and optimizing the driving experience. However, the existing vehicle-mounted atmosphere lamp control systems mostly adjust based on simple sensor data or preset lighting modes, lacking in-depth modeling between different driving scenarios, and it is difficult to accurately capture the complex changes and subtle features of the driving situation, resulting in the atmosphere lamp being unable to adapt to the dynamically changing driving environment in real time.

[0003] When traditional vehicle-mounted atmosphere lamp controllers adjust the lighting effects in different areas of the vehicle, there is generally a problem of color-brightness coupling, that is, when adjusting one parameter, it will inevitably affect other parameters, making precise control difficult. For example, adjusting the hue may cause an unexpected change in brightness, or increasing the brightness may affect the color saturation. This not only reduces the user experience but also limits the application effect of vehicle-mounted atmosphere lamps in specific driving scenarios. At the same time, the existing systems are difficult to provide differentiated lighting control for different driving areas (such as the driving area, the co-pilot area, and the rear row area), and cannot meet the personalized lighting needs of different occupants in different driving scenarios. Summary of the Invention

[0004] The present invention provides a vehicle-mounted atmosphere lamp control method and system based on environmental perception. The present invention solves the problem of unbalanced lighting requirements in different driving scenarios and makes the adjustment of vehicle-mounted atmosphere lamps more accurate.

[0005] In a first aspect, the present invention provides a vehicle-mounted atmosphere lamp control method based on environmental perception. The vehicle-mounted atmosphere lamp control method based on environmental perception includes:

[0006] Collect in-vehicle image data and vehicle-mounted sensor data and input them into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map;

[0007] Perform situation recognition and analysis on the fused feature map to obtain a driving situation feature vector and a situation correlation matrix;

[0008] Perform decoupling processing on the vehicle-mounted atmosphere lamp parameters according to the driving situation feature vector and the situation correlation matrix to obtain decoupled lighting parameters;

[0009] Perform situation - adaptive lighting control and dynamic situation - weighted loss calculation on the decoupled lighting parameters and the driving situation feature vector to obtain target lighting control parameters.

[0010] In a second aspect, the present invention provides an in - vehicle ambient light control system based on environmental perception. The in - vehicle ambient light control system based on environmental perception includes:

[0011] An acquisition module, configured to acquire in - vehicle image data and in - vehicle sensor data and input them into a dual - branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map;

[0012] A situation recognition module, configured to perform situation recognition and analysis on the fusion feature map to obtain a driving situation feature vector and a situation correlation matrix;

[0013] A decoupling processing module, configured to perform decoupling processing on the in - vehicle ambient light parameters according to the driving situation feature vector and the situation correlation matrix to obtain decoupled lighting parameters;

[0014] A loss calculation module, configured to perform situation - adaptive lighting control and dynamic situation - weighted loss calculation on the decoupled lighting parameters and the driving situation feature vector to obtain target lighting control parameters.

[0015] In a third aspect, a computer - readable storage medium is provided. Instructions are stored in the computer - readable storage medium. When it runs on a computer, it causes the computer to execute the above - mentioned in - vehicle ambient light control method based on environmental perception.

[0016] In the technical solution provided by the present invention, by adopting a dual-branch Mamba information fusion module to process in-vehicle image data and vehicle-mounted sensor data, effective fusion of multi-source heterogeneous data is achieved, and the perception ability of driving scenarios is improved. This module uses an image processing branch and a sensor data processing branch to extract spatial features and temporal features respectively, and performs adaptive fusion through a cross-modal feature fusion layer, overcoming the problem of insufficient information in a single data source. By using a multi-scale driving scenario attention module, multi-scale feature representations are extracted through convolutional kernels of different sizes, and self-attention mechanism and cross-scale attention fusion are applied to achieve accurate recognition and analysis of different driving scenarios, enabling the system to focus on the features most relevant to the adjustment of the ambient light. Aiming at the color-brightness coupling problem existing in traditional vehicle-mounted ambient light controllers, the present invention proposes a parameter decoupling processing mechanism. By establishing an LAFC parameter coupling relationship model and a double-input double-output disturbance observation mechanism, independent and accurate control of the four parameters of hue, saturation, lightness, and brightness is achieved, making the adjustment of vehicle-mounted ambient light more precise. Based on the context-adaptive lighting control method, the system can dynamically adjust lighting parameters according to the identified driving scenarios and perform differential control on different areas inside the vehicle, meeting the personalized lighting needs of different occupants in different driving scenarios. The dynamic context weighted loss function is innovatively introduced to solve the problem of unbalanced lighting requirements in different driving scenarios. By adjusting the context weight factor, the system can perform priority adjustment for specific driving scenarios and optimize the overall lighting effect. In addition, the system also continuously accumulates driving experience through the lighting strategy self-learning module and continuously improves the lighting control effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a schematic diagram of an embodiment of the vehicle-mounted ambient light control method based on environmental perception in the embodiment of the present invention;

[0019] Figure 2 It is a schematic diagram of an embodiment of the vehicle-mounted ambient light control system based on environmental perception in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] An embodiment of the present invention provides a vehicle ambient light control method and system based on environmental perception. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present invention are used to distinguish similar objects and do not necessarily need to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the term "comprising" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the specific process of the embodiment of the present invention will be described below. Please refer to Figure 1 , an embodiment of the vehicle ambient light control method based on environmental perception in the embodiment of the present invention includes:

[0022] Step S101, collect in-vehicle image data and vehicle-mounted sensor data and input them into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map;

[0023] It can be understood that the execution subject of the present invention can be a vehicle ambient light control system based on environmental perception, or a terminal or a server. Specifically, it is not limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example.

[0024] Specifically, a data acquisition system is installed in the vehicle. The high-resolution camera, as the main source of visual data, is installed at the interior rearview mirror to ensure comprehensive coverage of the driver's face and the main areas inside the vehicle. The acquisition frequency of this camera is set at 30 frames per second, and the resolution reaches 1920×1080 pixels to ensure the clarity and real-time nature of the acquired images. At the same time, an in-vehicle sensor array is carried, which includes a light sensor, a temperature sensor, an acceleration sensor, and a GPS positioning module. Among them, the light sensor is used to measure the light intensity inside and outside the vehicle, with a sampling frequency of 10 Hz and a measurement range covering 0 to 100,000 lux to ensure that light changes can be sensed in a timely manner; the temperature sensor collects the temperature data inside the vehicle, with its sampling frequency set at 1 Hz and a measurement accuracy reaching ±0.5°C to meet the requirements of in-vehicle environment monitoring; the acceleration sensor is used to detect the dynamic state of the vehicle, with a measurement range set at ±2g and a sampling frequency reaching 50 Hz to accurately capture the acceleration changes of the vehicle; the GPS module is responsible for obtaining the geographical location, speed, and driving direction data of the vehicle, with an update frequency set at 1 Hz to ensure that the vehicle's driving trajectory can be accurately recorded. The in-vehicle camera images and the data of the vehicle surrounding environment parameters are packed and processed according to a predefined data packet format. Each data packet contains a timestamp, a sensor ID, a data value, and a status flag. To improve the efficiency of data transmission and processing, the formatted data packets are classified according to different priorities, and all data are divided into real-time high-priority data streams and timed low-priority data streams. Among them, the real-time data stream contains the in-vehicle camera image data and the sensor data with relatively fast dynamic changes of the vehicle, such as acceleration information, while the low-priority data stream mainly contains the relatively slowly changing data such as environmental temperature and light intensity. On the basis of data stream stratification, the data stream is dynamically adjusted and processed in combination with the vehicle driving state to ensure that the system can collect the most critical information in different driving scenarios. For example, when the vehicle is stationary or driving at a low speed, the frame rate of image acquisition is reduced or the sampling frequency of some environmental parameters is decreased to reduce the computational burden, while in the case of high-speed driving or special driving situations, the data acquisition frequency is increased to ensure the integrity and timeliness of the data. The in-vehicle image data and the in-vehicle sensor data are input into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map.

[0025] Preprocess the in-vehicle image data and in-vehicle sensor data to ensure the consistency of data format and the optimization of computational efficiency. In the data acquisition stage, the in-vehicle camera records high-resolution images of 1920×1080 pixels at a frequency of 30 frames per second, while the in-vehicle sensor array, including a light sensor, a temperature sensor, an acceleration sensor, and a GPS module, provides environmental parameter data at sampling frequencies of 10Hz, 1Hz, 50Hz, and 1Hz respectively. After being acquired, these data will undergo normalization processing. Among them, the image data is cropped and scaled to a unified 256×256 pixel format to meet the input requirements of the neural network. Then, histogram equalization is performed to enhance the image contrast. At the same time, normalization processing is used to map the pixel values of the RGB channels to the interval [0,1], making the image features more standardized. For the in-vehicle sensor data, noise data is removed through outlier detection, and smoothing processing is performed using the moving window average method to ensure the continuity and stability of the data. After that, time series alignment is carried out to ensure that all sensor data has a unified timestamp, and after finally being converted into a standardized format, it is used for subsequent feature extraction. After completing the data preprocessing, the standardized images are input into the image processing branch of the dual-branch Mamba information fusion module. In this branch, initial feature extraction is performed through the first convolutional layer. This convolutional layer uses a 3×3 convolutional kernel, with a stride of 1, a padding of 1, and 64 channels, and ReLU is used as the activation function, enabling the effective extraction of low-level edge and texture features; the second convolutional layer continues to extract more complex local features on the basis of the previous layer, with the same parameter settings, but the number of channels is increased to 128 to enhance the feature representation ability; the third and fourth convolutional layers further extract high-level semantic features respectively, and a max pooling layer is added after each convolution, with the pooling kernel size set to 2×2 and the stride set to 2 to reduce the dimension of the feature map and retain the most representative key information. The final obtained image spatial features include key visual information such as the light distribution, color information, and driver status of the in-vehicle environment. At the same time, the standardized sensor data is input into the sensor data processing branch of the dual-branch Mamba information fusion module, and the Mamba state space model is used for time series feature extraction in this branch. The key parameters of the Mamba state space model are set as the state dimension E8, the unfolding step size 4, the expansion coefficient 2, and a bidirectional scanning method is adopted to ensure that the time-dependent relationship of the sensor data can be effectively captured. In the specific calculation process, this model performs state modeling on the input data, maps the sensor data to a high-dimensional state space to better describe its trend over time, and then performs sequence unfolding through a window with an unfolding step size of 4, enabling the system to consider the change information within adjacent time slices simultaneously, and using the expansion coefficient 2 to increase the data representation ability. At the same time, a bidirectional scanning mechanism is adopted, that is, not only predicting the current state based on past data, but also using future information for reverse adjustment to improve the integrity and accuracy of the time series features.Through this processing method, the system accurately extracts key dynamic features such as vehicle speed changes, light intensity fluctuations, temperature trends, and acceleration changes, forming sensor time-series features. The image spatial features and sensor time-series features are cross-modally fused to make full use of the complementarity of the two data types. These features are input into the cross-modal feature fusion layer of the dual-branch Mamba information fusion module. In this layer, the attention mechanism is used to calculate the attention weight matrix, the dimension of which is set to 128×128, to measure the importance relationship between different modal features. Subsequently, the Softmax function is used to normalize the weights, enabling the features of different modalities to adaptively adjust their influence during the fusion process to ensure the balanced fusion of information. Since image features are mainly used to describe spatial information, while sensor data is mainly used to describe time-evolution features, the attention mechanism can ensure that in different driving scenarios, the system pays more attention to the features that contribute most to the current state. For example, in the case of drastic light changes, it pays more attention to the light sensor data, and in the case of abnormal driver states, it pays more attention to the image information. To prevent the loss of original feature information during the fusion process and improve the training stability of the model, a residual connection is added to the adaptive fusion features to retain some original information, and layer normalization technology is used to prevent the shift of feature distributions, ensuring that the data can still maintain a stable numerical range during different batch training processes. The fused and optimized feature data is output as a fused feature map, the size of which is set to 128×64×64, containing multi-dimensional information such as in-vehicle visual information, environmental parameters, and vehicle dynamic states.

[0026] Step S102: Perform situation recognition and analysis on the fused feature map to obtain a driving situation feature vector and a situation correlation matrix;

[0027] Specifically, the fused feature map passes through a multi-scale feature extraction module. This module performs convolution operations using three different-sized convolutional kernels, namely convolutional kernels with sizes of 3×3, 5×5, and 7×7 respectively, to ensure that feature information at different scales can be captured. The small-sized convolutional kernel extracts local detail information such as edges and textures, while the medium-sized convolutional kernel is used to identify medium-range feature patterns, such as object contours and lighting distributions. The larger-sized convolutional kernel helps to extract global information, including the overall brightness trend and spatial structure. Through the extraction process of the three different-sized convolutional kernels, three feature representations at different scales are obtained. To enhance the correlation between features and mine the internal patterns of the data, the self-attention mechanism is applied to these three feature representations at different scales, and the scaled dot-product attention algorithm is used to calculate the spatial correlation within the feature map. This algorithm obtains query, key, and value matrices through linear transformation, then calculates the dot product of the query and key matrices, and uses the square root of the feature dimension as the scaling factor to ensure numerical stability. Then, the Softmax function is applied to the calculated attention weights to normalize them, so as to ensure that the sum of all attention distributions is 1, and the value matrix is weighted and summed using the normalized weights to generate an attention-weighted feature map. Since feature maps at different scales have different spatial distribution characteristics, the scaled dot-product attention algorithm can ensure that the system can find the most important feature regions at different scales, thereby improving the recognition ability of driving scenarios and generating three attention-weighted feature maps at different scales. The three attention-weighted feature maps at different scales are input into the cross-scale attention fusion module to achieve adaptive fusion of information. During this process, learnable weight parameters are introduced, enabling the fusion process to automatically adjust the contribution ratio of features at different scales to ensure the optimality of the final fused features. The cross-scale attention fusion module maps feature vectors at different scales to the same feature space, and then weights and sums them using the weight parameters, and these weight parameters will be continuously adjusted during the training process to minimize the error of driving scenario classification, so that the fused features can fully express multi-scale information and obtain a fused feature vector. The fused feature vector is processed through a fully connected layer to output the driving scenario category distribution. At this stage, the fused feature vector is fed into a fully connected neural network, which contains multiple hidden layers and uses non-linear activation functions to enhance the non-linear expression ability of the data. The Softmax function is used in the output layer to map the feature vector to the probability distribution of driving scenario categories. This probability distribution represents the classification possibilities of different driving scenarios, such as highway cruising, urban congestion, night driving, rainy day driving, hard braking, etc., and the current driving scenario is determined according to the highest probability category.After obtaining the driving situation category, key feature indicators related to the situation are extracted to generate a driving situation feature vector. The feature vector contains four key indicators, including lighting conditions, driver status, vehicle speed change rate, and environmental complexity. Among them, lighting conditions are measured by an in-vehicle light sensor to quantify the brightness level inside the vehicle. Driver status is analyzed by an in-vehicle camera to observe the driver's facial expression and attention level to judge the driver's mental state. The vehicle speed change rate is measured by GPS and an acceleration sensor to reflect the vehicle's driving mode. Environmental complexity is comprehensively evaluated by combining vehicle speed changes, external lighting conditions, and driver status to judge the stability of the current driving scenario. Through these four key features, the current driving situation is characterized. Based on the driving situation feature vector, the correlation coefficients between different driving situations and the ambient light parameters are calculated to generate a situation correlation matrix. This matrix is used to quantify the weights of the ambient light adjustment requirements in different situations. Its calculation method is based on statistical analysis and regression analysis of historical data. The size of this matrix is E×4, where E represents the number of driving situations, and 4 corresponds to the four parameters of the ambient light, including hue, saturation, lightness, and brightness. To calculate the correlation coefficients, a large amount of historical data is collected, and the change trends of each ambient light parameter in each driving situation are evaluated through a regression analysis model. Then, by calculating the Pearson correlation coefficient, the dependence relationships between these parameters are quantified, and finally a complete situation correlation matrix is formed. Each element of this matrix represents the influence degree of a certain driving situation on a certain ambient light parameter and is used to guide subsequent adaptive lighting control.

[0028] Step S103: Decouple the in-vehicle ambient light parameters according to the driving situation feature vector and the situation correlation matrix to obtain the decoupled lighting parameters;

[0029] Specifically, by experimentally measuring the visual effects under different RGB color combinations and brightness levels, a coupled relationship model of LAFC parameters is constructed. During the experiment, a standard lighting environment is adopted to analyze the color visual perception under different RGB ratio combinations. By combining subjective experiments and objective measurements, the influence degrees of different lighting conditions on hue, saturation, lightness, and brightness are recorded to form a coupling coefficient matrix, which is used to describe the mutual influence relationship among the four parameters, enabling the system to quantitatively evaluate the influence degree of a change in one parameter on other parameters. Based on the coupling coefficient matrix, the HSV color space is divided into multiple grids, and the corresponding relationship between the visual effect and the RGB brightness is calculated at each grid point to form a four-dimensional parameter mapping table. The grid division method of this mapping table is based on the statistical results of experimental data. The hue, saturation, and lightness are divided with a resolution of 36×10×10, and a complete four-dimensional parameter mapping table is constructed with a 100-level brightness distribution. The change range of the hue is 0 - 359 degrees, divided at 10-degree intervals, the change ranges of saturation and lightness are 0 - 100%, divided at 10% intervals, and the brightness parameter is finely divided at 1% intervals. Based on the four-dimensional parameter mapping table, a dual-input and dual-output disturbance observation mechanism is designed. This mechanism measures the influence relationship among parameters through a small perturbation experiment and performs a preset perturbation within the error range to observe the influence of different parameter changes on the overall visual effect. Based on the current parameter settings, the system applies a ±5% small perturbation to the values of hue, saturation, lightness, and brightness and records the visual effect change data before and after the perturbation. These data can help the system quantify the interactive influence among parameters. According to the visual effect change data before and after the perturbation, the local Jacobian matrix between parameters is calculated. This matrix is used to describe the sensitivity relationship of each lighting parameter near the current operating point. By means of differential approximation, the change amounts of different parameters are compared with the change amount of the visual effect to form a parameter sensitivity matrix. The size of this matrix is 4×4, and each element represents the partial derivative relationship among hue, saturation, lightness, and brightness, reflecting the influence degree of each parameter on the overall visual effect during a small adjustment. A parameter compensator is constructed based on the inverse matrix of the parameter sensitivity matrix. This compensator cancels the cross influence among different parameters through matrix transformation to achieve independent control. For this purpose, the inverse matrix of the parameter sensitivity matrix is calculated, and a decoupling transformation matrix is constructed using this inverse matrix. The numerical range of this transformation matrix is limited between [-0.5, 0.5] to ensure the smoothness of the adjustment process and prevent sudden changes in the visual effect caused by excessive parameter changes. The initial lighting parameters are determined according to the driving scenario feature vector and the scenario correlation matrix, and are adjusted using the decoupling transformation matrix to ensure that the finally output lighting parameters can adapt to the current driving environment.According to the driving scenario category, extract the corresponding weight for adjusting the ambient light parameters from the scenario correlation matrix, and combine information such as the lighting conditions, driver status, vehicle speed change rate, and environmental complexity in the driving scenario feature vector to determine the setting values of the initial lighting parameters. These setting values include the default values of the basic hue, saturation, lightness, and brightness, and are dynamically adjusted in combination with the scenario information to obtain the initial lighting parameters. Input the initial lighting parameters into the decoupling transformation matrix for processing to automatically calculate and adjust the mutual influence between the parameters and obtain the final decoupled lighting parameters.

[0030] Step S104: Perform scenario adaptive lighting regulation and dynamic scenario weighted loss calculation on the decoupled lighting parameters and the driving scenario feature vector to obtain the target lighting control parameters.

[0031] Specifically, a situation-illumination mapping database is constructed for E typical driving scenarios. This database contains the ideal lighting parameters of in-vehicle ambient lights in different driving scenarios and defines the recommended values and adjustable ranges of four basic parameters: hue, saturation, lightness, and brightness. The setting of these parameters is based on long-term experimental measurements and user feedback data, and comprehensively considers driving safety, comfort, and the psychological feelings of the driver. Fuzzy logic control is performed according to the current driving scenario feature vector to calculate the initial lighting parameter adjustment amount. The input variables of fuzzy logic control include driving scenario category, scenario confidence, vehicle speed, ambient light intensity, and driver state, while the output variables include hue adjustment amount, saturation adjustment amount, lightness adjustment amount, and brightness adjustment amount. Each input variable is divided into multiple fuzzy sets. For example, vehicle speed is divided into low speed, medium speed, and high speed, and ambient light intensity is divided into low light, normal light, and strong light, and a series of fuzzy rules are set to deduce the output. After calculating the initial lighting parameter adjustment amount, optimize this parameter to ensure the smoothness of light changes. Therefore, dynamic optimization processing is performed on the initial adjustment amount based on the real-time monitored driving scenario change rate. When the driving scenario change rate is small, a smaller adjustment step is adopted to make the change of the ambient light smoother, while when the driving scenario change rate is large, a larger adjustment step is adopted to ensure that the lighting can quickly adapt to the new driving scenario. The adaptive optimization strategy can effectively improve the stability of lighting regulation and make the change of the ambient light more in line with the physiological and psychological perception of the driver. To optimize lighting regulation, the interior space of the vehicle is divided into 4 independent lighting areas, including the driving area, co-driver area, left rear row, and right rear row, and the calculated smooth adjustment parameters are processed with regional differences to meet the needs of occupants in different areas. Specifically, according to the activity areas and positions of the driver and passengers, the lighting parameters are independently adjusted for different areas. After completing the calculation of the regional lighting control parameters, match them with the lighting template parameters in the situation-illumination mapping database to evaluate the closeness of the current lighting scheme to the preset best lighting scheme. Calculate the Euclidean distance between the regional lighting control parameters and the lighting template parameters, and evaluate the lighting effect matching degree according to the calculation result. The calculation method of the Euclidean distance is to normalize the four parameters of hue, saturation, lightness, and brightness to the same scale range and calculate their distance in the four-dimensional space. The higher the matching degree, the closer the current lighting scheme is to the best setting. When the lighting effect matching degree is lower than the preset target score, the system will automatically enter the optimization mode to further fine-tune the lighting parameters.During the optimization process, a binary search method is adopted to finely tune the regional lighting control parameters, calculate the adjustment direction of the current lighting parameters, and attempt to find a better parameter combination in the parameter space. Each time an adjustment is made, the new Euclidean distance is calculated, and it is determined whether the matching degree has improved. The system will continuously execute this process until the matching degree reaches the preset threshold or the number of adjustments reaches the maximum allowed value. With the support of this optimization strategy, the system can ensure that the final lighting parameters are as close as possible to the optimal settings, so as to improve the adaptability and consistency of the in-vehicle lighting experience. After optimizing the lighting parameters, the dynamic scenario weighted loss is calculated to ensure that the final target lighting control parameters can adapt to the specific requirements of different driving scenarios. A driving experience evaluation model is constructed, and the current lighting effect is scored according to four dimensions: visual comfort, driving assistance, mood regulation, and scenario matching degree. The weight of each dimension is initially set to 0.25 and dynamically adjusted based on the driver's facial expressions, eye movements, and operating behaviors captured by the in-vehicle camera. The scoring range is set between 0 and 100 to reflect the adaptability of the lighting scheme in different driving scenarios. The dynamic scenario weighted loss function is used to calculate the final target lighting control parameters, and this loss function is weighted by the importance of different driving scenarios to optimize the lighting control strategy. The gradient descent optimization algorithm is used to iteratively calculate the loss function and adjust the lighting parameters to make them as close as possible to the target settings. To ensure the stability of the optimization process, the learning rate is set to 0.01, the maximum number of iterations is 100, and the convergence threshold is set to 0.001 to ensure that the loss function can converge within a reasonable range. The target lighting control parameters are calculated and transmitted to the in-vehicle ambient light hardware execution unit, enabling it to automatically adjust the lighting scheme in different driving scenarios, thereby enhancing the driving experience and improving driving safety.

[0032] An evaluation model is constructed based on the driver's facial expressions, eye movements, and operating behaviors. The model uses an in-vehicle camera to capture the driver's facial expressions in real time and combines deep learning algorithms to classify the expression features, thereby judging the driver's emotional state. At the same time, eye-tracking technology is used to analyze the driver's line-of-sight direction and blink frequency to detect the attention level and potential fatigue state. In addition, the vehicle control unit monitors the driver's operating behaviors, such as steering angle, braking force, and throttle control method, to evaluate the driver's driving style and behavior pattern. After these data are fused, they are used to calculate four key evaluation dimensions, namely visual comfort, driving assistiveness, emotion regulation, and context matching. Among them, visual comfort measures whether the current lighting environment interferes with the driver's visual perception; driving assistiveness evaluates whether the lighting conditions help improve driving safety; emotion regulation measures the impact of the lighting scheme on the driver's mental state; and context matching reflects whether the current lighting meets the requirements of the expected driving scenario. The scores of these four dimensions together constitute the experience score, which is set in the range of 0 to 100 to reflect the adaptability of different lighting parameters and the user experience. After obtaining the experience score, a dynamic context weighted loss function is applied to calculate the initial lighting control parameters to quantify the deviation between the lighting parameters and the best experience. This loss function is weighted based on the importance of the driving context to optimize the lighting control strategy. Based on the calculated initial loss value, a gradient descent optimization algorithm is executed to iteratively optimize the lighting parameters. The learning rate is set to 0.01, the maximum number of iterations is set to 100, and the convergence threshold is set to 0.001. At each iteration, the partial derivative of the loss function with respect to the lighting parameters is calculated, and the parameter values are adjusted according to the direction of gradient descent to minimize the loss, ensuring that the finally optimized lighting parameters can best fit the optimal lighting scheme, thereby enhancing the driving experience and safety. After the optimization is completed, a safety constraint test is performed on the optimized lighting parameters to ensure that they meet the road safety standards and the requirements of human eye visual comfort. This test process includes brightness limit, color temperature regulation, and hue change constraint. Among them, the brightness limit ensures that the light intensity of the ambient light does not exceed the maximum brightness specified by the regulations to avoid affecting the line of sight of the driver or other road users. The color temperature regulation ensures that the lighting scheme is within the range of human eye visual comfort to avoid discomfort caused by too cold or too warm color temperature. At the same time, the hue change constraint ensures that the hue adjustment rate of the ambient light does not exceed the preset threshold to avoid sudden color changes affecting the driver's attention. The lighting parameters that pass this safety test are regarded as lighting parameters that meet the safety standards and are used for the final control execution.Compare the lighting parameters that meet safety standards with the initial lighting control parameters through a loss function evaluation to quantify the actual effect of the optimization process. This evaluation method calculates the difference in loss values before and after optimization to evaluate the improvement amplitude of the optimization strategy. If the loss value after optimization decreases significantly, it indicates that the optimization strategy effectively improves the adaptability of the lighting scheme. If the change in the loss value is not significant, it means that the current optimization strategy needs further adjustment. Calculate the difference in experience scores before and after optimization to verify whether the optimization improves the actual experience of the driver, and evaluate the impact of different lighting parameters on the driving experience through statistical analysis. All these evaluation data will be stored as optimization effect data. To ensure that the system can continuously improve and adapt to changes in different driving environments, based on the optimization effect data and experience scores, use an incremental learning method to update the situation-lighting mapping database. This update process adopts a batch update method, and a parameter adjustment is performed every time 100 driving data are accumulated. During the update process, combine the newly collected data, recalculate the optimal lighting parameters in different situations, and update the lighting template. At the same time, during the incremental learning process, use historical data for regression analysis to ensure that the updated database will not be over-adjusted due to short-term data deviations, thus maintaining long-term stability and adaptability. Output the optimized target lighting control parameters and transmit them to the in-vehicle ambient light hardware execution unit through the CAN bus to achieve precise lighting control, so as to ensure that the ambient light can always provide the best visual experience and safety guarantee in different driving situations.

[0033] In an embodiment of the present invention, by adopting a dual-branch Mamba information fusion module to process in-vehicle image data and vehicle-mounted sensor data, effective fusion of multi-source heterogeneous data is achieved, and the perception ability of the driving situation is improved. This module uses an image processing branch and a sensor data processing branch to extract spatial features and temporal features respectively, and performs adaptive fusion through a cross-modal feature fusion layer, overcoming the problem of insufficient information in a single data source. A multi-scale driving situation attention module is adopted to extract multi-scale feature representations through convolutional kernels of different sizes, and apply self-attention mechanism and cross-scale attention fusion to achieve accurate recognition and analysis of different driving situations, enabling the system to focus on the features most relevant to the adjustment of the ambient light. Aiming at the color-brightness coupling problem existing in traditional vehicle-mounted ambient light controllers, the present invention proposes a parameter decoupling processing mechanism. By establishing a LAFC parameter coupling relationship model and a double-input double-output disturbance observation mechanism, independent and accurate control of the four parameters of hue, saturation, lightness, and brightness is achieved, making the adjustment of vehicle-mounted ambient light more precise. Based on the situation-adaptive lighting control method, the system can dynamically adjust lighting parameters according to the recognized driving situation and perform differential control on different areas inside the vehicle, meeting the personalized lighting needs of different occupants in different driving scenarios. The dynamic situation weighted loss function is innovatively introduced to solve the problem of unbalanced lighting requirements in different driving scenarios. By adjusting the situation weight factor, the system can perform priority adjustment for specific driving situations and optimize the overall lighting effect. In addition, the system also continuously accumulates driving experience through the lighting strategy self-learning module and continuously improves the lighting control effect.

[0034] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0035] Collect images of the in-vehicle environment through a high-resolution camera installed at the in-vehicle rearview mirror to obtain in-vehicle camera images;

[0036] Collect environmental parameters using a vehicle-mounted sensor array, which includes a light sensor, a temperature sensor, an acceleration sensor, and a GPS positioning module, to obtain vehicle surrounding environmental parameter data;

[0037] Pack the in-vehicle camera images and the vehicle surrounding environmental parameter data according to a predefined data packet format to obtain a formatted data packet. The predefined data packet format includes a timestamp, a sensor ID, a data value, and a status flag;

[0038] Divide the formatted data packet into a real-time high-priority data stream and a timed low-priority data stream to obtain a hierarchical data stream, and perform dynamic adjustment processing on the hierarchical data stream according to the vehicle driving state to obtain in-vehicle image data and vehicle-mounted sensor data;

[0039] Input the in-vehicle image data and vehicle-mounted sensor data into the dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map.

[0040] Specifically, install a high-resolution camera at the in-vehicle rearview mirror. This camera captures high-definition images of 1920×1080 pixels at a rate of 30 frames per second, ensuring full coverage of the driver's face and the main areas inside the vehicle to obtain accurate in-vehicle environment images. In cooperation with the camera, use a vehicle-mounted sensor array to collect environmental parameters. This sensor array includes a light sensor, a temperature sensor, an acceleration sensor, and a GPS positioning module. Among them, the light sensor is used to measure the light intensity inside and outside the vehicle. Its sampling frequency is set to 10Hz, and the measurement range is from 0 to 100,000 lux, ensuring that it can accurately sense the light changes in the in-vehicle environment. The temperature sensor is used to record the in-vehicle temperature. Its accuracy reaches ±0.5°C, and the sampling frequency is set to 1Hz to ensure the real-time nature of the data. The acceleration sensor is used to measure the dynamic acceleration of the vehicle. The range is set to ±2g, and the sampling frequency reaches 50Hz, thus ensuring that it can accurately capture the acceleration changes of the vehicle. The GPS module, on the other hand, obtains the vehicle's position information, driving speed, and direction in real time. Its update frequency is 1Hz to ensure the accuracy of the positioning information. Format the in-vehicle camera images and vehicle surrounding environmental parameter data to ensure efficient storage and calculation of data from different sources. All data is uniformly packaged according to a predefined data packet format. Each data packet contains a timestamp, a sensor ID, a data value, and a status flag. This formatting method helps to standardize the storage of data and ensures accurate alignment of data during the multi-source information fusion process, thereby reducing the possibility of information loss. To improve the efficiency of data transmission and calculation, divide the formatted data packets into two types of data streams, namely real-time high-priority data streams and timed low-priority data streams. Among them, the real-time high-priority data streams mainly include in-vehicle camera image data and sensor data with relatively fast vehicle dynamics changes, such as acceleration information. The timed low-priority data streams include environmental parameter data with relatively slow changes, such as environmental temperature and light intensity. Based on the division of the data streams, dynamically adjust the data streams in combination with the vehicle driving state to ensure the most critical information is collected in different driving scenarios. For example, when the vehicle is driving at a low speed or stationary, reduce the frame rate of image acquisition and the sampling frequency of some sensors to reduce the computational burden. In the case of high-speed driving or complex driving states, increase the data acquisition frequency to ensure the integrity and timeliness of the data. And set a dynamic adjustment strategy based on adaptive information entropy. This strategy measures the uncertainty of the current information by calculating the data entropy and adjusts the data acquisition frequency according to the change trend of the uncertainty. Specifically, define the data entropy as:

[0041]

[0042] Among them, represents the normalized probability of the th class of data, represents the total number of data categories. When the data entropy is high, it indicates that the environment changes greatly, and the system automatically increases the data sampling frequency to obtain more information. When the data entropy is low, it means that the environment is relatively stable, and the system appropriately reduces the data sampling frequency to reduce the computational burden and improve the processing efficiency. After completing the dynamic adjustment of the data stream, the obtained in-vehicle image data and vehicle-mounted sensor data are input into the dual-branch Mamba information fusion module for feature extraction and fusion processing. In this module, the data is respectively sent into two independent processing branches. Among them, the image data enters the image processing branch, and this branch uses a multi-layer convolutional neural network for feature extraction. The convolutional kernel size of the first convolutional layer is set to 3×3, the stride is 1, the number of channels is set to 64, and ReLU is used as the activation function to extract the basic edge features. Subsequently, the number of channels of the second convolutional layer is increased to 128 to extract higher-level image semantic information. After each convolutional layer, there is a max-pooling layer, and the pooling kernel size is 2×2, and the stride is set to 2 to reduce the computational complexity and retain the most critical feature information. The sensor data enters the Mamba state space model, which is specifically used to capture time series features. In this model, all sensor data are time-aligned, and a bidirectional scanning method with a state dimension of 128, an unfolding stride of 4, and an expansion coefficient of 2 is adopted to ensure that time series information can be effectively extracted, so as to accurately model the dynamic state of the vehicle and the environmental change trend. After completing the feature extraction of the two data branches, cross-modal fusion is performed on the two types of features to make full use of the complementarity of image and sensor data. These features are input into the cross-modal feature fusion layer, which uses an attention mechanism to calculate the importance weights of the features and normalizes the weights through the Softmax function, so that the features of different modalities can adaptively adjust their influence during the fusion process, thereby ensuring that the final fused feature map can fully express the key feature information of the current driving situation. To optimize the feature fusion effect, an optimization strategy based on feature mutual information is introduced, and this strategy defines the feature mutual information as:

[0043]

[0044] Among them, and respectively represent the feature data of two different modalities, represents the joint probability distribution, and Respectively representing their respective marginal probability distributions, by maximizing the feature mutual information, the system ensures that the fused features retain as much key information of the image and sensor data as possible without losing important details. The output fused feature map contains visual information about the in-vehicle environment while integrating in-vehicle sensor data, thus comprehensively reflecting the driving state and environmental changes of the vehicle.

[0045] In a specific embodiment, the process of inputting in-vehicle image data and in-vehicle sensor data into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map may specifically include the following steps:

[0046] Perform image preprocessing on the in-vehicle image data to obtain a standardized image, and perform data preprocessing on the in-vehicle sensor data to obtain standardized sensor data;

[0047] Input the standardized image into the image processing branch of the dual-branch Mamba information fusion module, and extract spatial features through 4 convolutional layers to obtain image spatial features;

[0048] Input the standardized sensor data into the sensor data processing branch of the dual-branch Mamba information fusion module, and extract temporal features through the Mamba state space model. The Mamba state space model is set with a state dimension of E8, an unfolding step size of 4, an expansion coefficient of 2, and a bidirectional scanning method to obtain sensor temporal features;

[0049] Input the image spatial features and sensor temporal features into the cross-modal feature fusion layer of the dual-branch Mamba information fusion module, calculate the attention weight matrix using the attention mechanism and normalize it through the Softmax function to obtain adaptive fusion features;

[0050] Add a residual connection to the adaptive fusion features to retain the original feature information, and use layer normalization to prevent the feature distribution from shifting to obtain a fused feature map.

[0051] Specifically, image preprocessing is performed on the in-vehicle image data. The image is cropped to focus on the driver's face area and the main in-vehicle environment area, ensuring the integrity of key information and minimizing background noise. The image is scaled to 256×256 pixels to reduce the computational burden and adapt to subsequent neural network processing. Histogram equalization technology is applied during the scaling to enhance the image contrast and reduce the impact of light changes. At the same time, to eliminate the bias caused by uneven illumination, the RGB channels of the image are normalized, mapping the pixel values to the [0,1] interval, improving the stability of the model, and preventing the feature distribution drift caused by different lighting conditions. While performing image data preprocessing, data preprocessing is also carried out on the vehicle-mounted sensor data to ensure that data from different types of sensors can be fused and analyzed under the same time reference. Since different vehicle-mounted sensors have different sampling frequencies and data formats, time series alignment is performed, and interpolation is used to complete the data according to the timestamps of each sensor to ensure that all data points can correspond to the same time step. Outlier detection is performed on the sensor data. For example, for the light sensor, if the changes in multiple consecutive data points exceed the set physical limit, these points are marked as outliers and smoothed using the moving window average method to reduce the impact of environmental noise on the data. All sensor data is standardized to have a mean of zero and a variance of one, thus ensuring the consistency of the data distribution and preventing the uneven impact of the differences in the data ranges of different sensors on model learning. After completing the data preprocessing, the standardized image is input into the image processing branch of the dual-branch Mamba information fusion module. This branch uses a four-layer convolutional neural network for spatial feature extraction. The convolutional kernel size of the first convolutional layer is set to 3×3, the stride is 1, and the number of channels is 64. The ReLU function is used as the activation function to extract basic edge features. The number of channels in the second convolutional layer is increased to 128, and higher-level local pattern features are continuously extracted. The third and fourth convolutional layers further extract more complex spatial features, and a max pooling layer is added after each convolutional layer. The pooling kernel size is 2×2, and the stride is set to 2 to reduce the computational complexity and retain the most representative spatial information. After four-layer convolutional processing, image spatial features containing in-vehicle environment information, light change features, and driver status are extracted. At the same time, the standardized sensor data is input into the sensor data processing branch of the dual-branch Mamba information fusion module. This branch uses the Mamba state space model for temporal feature extraction, where the state dimension is set to E8, the unfolding step is 4, the expansion coefficient is 2, and a two-way scanning method is used to ensure that the time dependence of sensor data can be effectively captured.The model models the state of the input data, maps the sensor data to a high-dimensional state space to better describe its trend over time. Subsequently, sequence unfolding is performed by expanding a window with a step size of 4, enabling the system to consider the change information within adjacent time slices simultaneously. The expansion coefficient 2 is used to increase the data representation ability. Meanwhile, a bidirectional scanning mechanism is adopted, that is, not only predicting the current state based on past data, but also using future information for backward adjustment to improve the integrity and accuracy of time series features. To enhance the temporal modeling ability of the Mamba state space model, an optimization method based on an adaptive memory gating mechanism is defined. This method adjusts the weights of past time steps to adapt to data change patterns in different environments. This mechanism ensures that the system adaptively adjusts the degree of attention to different historical data and still maintains a stable temporal modeling ability in complex driving environments. After extracting the image space features and sensor temporal features, these two types of features are input into the cross-modal feature fusion layer of the dual-branch Mamba information fusion module. This layer uses an attention mechanism to calculate the importance weights of the features and normalizes them through the Softmax function, enabling the features of different modalities to adaptively adjust their influence during the fusion process, thus ensuring that the final fused features can fully represent the key information of the current driving situation. To prevent the loss of original feature information during the fusion process and improve the training stability of the model, a residual connection is added to the adaptive fused features to retain some original information, and layer normalization technology is used to prevent the shift of feature distributions, ensuring that the data can still maintain a stable numerical range during different batch training processes. The feature data after fusion and optimization is output as a fused feature map, whose size is set to 128×64×64, containing multi-dimensional information such as in-vehicle visual information, environmental parameters, and vehicle dynamic states.

[0052] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0053] Feature extraction is performed on the fused feature map through three convolutional kernels of different sizes to obtain three feature representations of different scales;

[0054] The self-attention mechanism is applied to the three feature representations of different scales respectively, and the scaled dot-product attention algorithm is used to calculate the spatial correlation within the feature map, obtaining three attention-weighted feature maps of different scales;

[0055] The three attention-weighted feature maps of different scales are input into the cross-scale attention fusion module and adaptively fused through learnable weight parameters to obtain a fused feature vector;

[0056] Fully connected processing is performed on the fused feature vector to output the driving situation category distribution;

[0057] Based on the driving scenario category distribution, extract the target feature indicators for each driving scenario to obtain a driving scenario feature vector. The target feature indicators include lighting conditions, driver status, vehicle speed change rate, and environmental complexity;

[0058] Conduct a quantitative analysis on the driving scenario feature vector, calculate the correlation coefficients between E driving scenarios and 4 kinds of ambient light parameters, and generate a scenario correlation matrix. The 4 kinds of ambient light parameters include hue, saturation, lightness, and brightness.

[0059] Specifically, multi-scale feature extraction is performed on the fused feature map to capture environmental information at different scales. Three different-sized convolutional kernels are used to perform convolutional operations on the fused feature map, with convolutional kernels of 3×3, 5×5, and 7×7 respectively, to ensure that local detail features, medium-scale structural information, and global patterns can be extracted. The small-sized convolutional kernel focuses on extracting local edge information, such as the boundaries of in-vehicle objects, the details of the driver's eyes, or the brightness distribution of the dashboard, while the medium-sized convolutional kernel is used to identify medium-scale features such as illuminated areas and cockpit layouts, and the larger-sized convolutional kernel extracts overall spatial relationships, such as the global distribution of in-vehicle lighting and the overall sitting posture characteristics of the driver. Through this multi-scale convolution strategy, three feature representations at different scales are obtained. After obtaining the feature representations at different scales, the spatial correlation within the feature map is calculated. Therefore, the self-attention mechanism is applied to these three feature representations respectively, and the scaled dot-product attention algorithm is used for calculation. This algorithm transforms the input features into queries, keys, and values, and calculates the dot product of the query and key matrices to measure the similarity between features. To ensure numerical stability, it is scaled using the square root of the feature dimension. Subsequently, the Softmax function is applied to the calculated attention weights for normalization, so that the sum of the attention distributions is 1. Finally, the value matrix is weighted and summed using the normalized attention weights to generate an attention-weighted feature map. Since the feature maps at different scales have different spatial distribution characteristics, the scaled dot-product attention algorithm can ensure that the system can find the most important feature regions at different scales, thereby improving the recognition ability of driving scenarios, generating three attention-weighted feature maps at different scales, and each feature map contains feature information redistributed based on the attention mechanism, making it more in line with the actual characteristics of driving scenarios. The three attention-weighted feature maps at different scales are input into the cross-scale attention fusion module. In this module, learnable weight parameters are introduced to enable the fusion process to automatically adjust the contribution ratios of features at different scales to ensure the optimality of the final fused features. The cross-scale attention fusion module maps the feature vectors at different scales to the same feature space, and then uses the weight parameters to perform weighted summation on them, and these weight parameters will be continuously adjusted during the training process to minimize the error of driving scenario classification, so that the fused features can fully express multi-scale information, thus obtaining a fused feature vector containing rich environmental information. The fused feature vector is subjected to a fully connected process to output the driving scenario category distribution. At this stage, the fused feature vector is fed into a fully connected neural network, which contains multiple hidden layers and uses non-linear activation functions to enhance the non-linear expression ability of the data. The Softmax function is used in the output layer to map the feature vector to the probability distribution of driving scenario categories, and this probability distribution represents the classification possibilities of different driving scenarios, such as highway cruising, urban congestion, night driving, rainy day driving, hard braking, etc., and the current driving scenario is determined according to the highest probability category.After obtaining the driving situation category, key feature indicators related to the situation are extracted to generate a driving situation feature vector. This feature vector contains four key indicators, including lighting conditions, driver status, vehicle speed change rate, and environmental complexity. Among them, lighting conditions are measured by an in-vehicle light sensor to quantify the brightness level inside the vehicle. Driver status is analyzed by an in-vehicle camera to observe the driver's facial expressions and attention level to judge the driver's mental state. The vehicle speed change rate is measured by GPS and an acceleration sensor to reflect the vehicle's driving mode. Environmental complexity is comprehensively evaluated by combining vehicle speed changes, external lighting conditions, and driver status to judge the stability of the current driving scenario. Through these four key features, the current driving situation is accurately characterized. Based on this driving situation feature vector, the correlation coefficient between different driving situations and the ambient light parameters is calculated to generate a situation correlation matrix. This matrix is used to quantify the weights of the ambient light adjustment requirements in different situations. Its calculation method is based on statistical analysis and regression analysis of historical data. The size of this matrix is E×4, where E represents the number of driving situations, and 4 corresponds to the four parameters of the ambient light, including hue, saturation, brightness, and luminance. To calculate the correlation coefficient, a large amount of historical data is collected, and the change trend of each ambient light parameter in each driving situation is evaluated through a regression analysis model. Then, the Pearson correlation coefficient is calculated to quantify the dependence relationship between these parameters. The calculation formula is as follows:

[0060]

[0061] Where, represents the correlation coefficient between the th driving situation and the th ambient light parameter, represents the driving situation feature value in the th observation, is the mean value of this situation feature value, represents the ambient light parameter value in the th observation, is the mean value of this ambient light parameter value, represents the sample size. By calculating the correlation coefficient, a situation correlation matrix is obtained. Each element of this matrix represents the influence degree of a certain driving situation on a certain ambient light parameter.

[0062] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0063] Measure the visual effects under different RGB color combinations and brightness levels through experiments, establish a coupling relationship model of LAFC parameters, and obtain a coupling coefficient matrix. The coupling coefficient matrix represents the mutual influence degree between the four parameters of hue, saturation, brightness, and luminance;

[0064] The HSV color space is divided into multiple grids based on the coupling coefficient matrix, and the corresponding relationship between the visual effect and the RGB brightness is calculated for each grid to obtain a four-dimensional parameter mapping table;

[0065] Based on the four-dimensional parameter mapping table, small perturbations within a preset error range are performed to construct a double-input and double-output perturbation observation mechanism, and the visual effect change data before and after the perturbation are obtained;

[0066] According to the visual effect change data before and after the perturbation, the local Jacobian matrix between the parameters is calculated to obtain the parameter sensitivity matrix, and a parameter compensator is constructed based on the inverse matrix of the parameter sensitivity matrix to obtain the decoupling transformation matrix;

[0067] Based on the driving scenario feature vector and the scenario correlation matrix, the initial lighting parameters are determined, and the initial lighting parameters are input into the decoupling transformation matrix for processing. By automatically calculating and adjusting the mutual influence between the parameters, the decoupled lighting parameters are obtained.

[0068] Specifically, the visual effects under different RGB color combinations and brightness levels are measured through experiments to establish an LAFC parameter coupling relationship model. The core of this model lies in quantitatively analyzing the mutual influence among the four parameters of hue, saturation, lightness, and brightness. During the experiment, a high-precision photometer and a standardized test environment are used to record the color perception data under different RGB combinations, and through psychophysical experiments, the subjective evaluations of users on different lighting conditions are collected. At the same time, objective data are obtained by combining a spectral analyzer to quantify the comprehensive influence of different lighting parameters on human eye visual perception. Through regression analysis of a large amount of experimental data, a coupling coefficient matrix is constructed, which is used to describe the dependence relationship among hue, saturation, lightness, and brightness. For example, under a specific hue, increasing the brightness will cause the saturation to decrease, and certain specific lightness changes will cause the hue to shift. These non-linear coupling relationships are accurately quantified for subsequent decoupling calculations. The HSV color space is divided into grids based on the coupling coefficient matrix to construct a color mapping table. By discretizing the HSV space, multiple high-dimensional data points are formed. Each grid point represents a fixed combination of hue, saturation, and lightness, and the corresponding relationship between the visual effect and the RGB brightness is calculated at each grid point. The HSV color space is divided into grids, where is the number of divisions of the hue, is the number of divisions of the saturation, is the number of divisions of lightness. The RGB brightness at each grid point is fitted with the data measured in the experiment to ensure an accurate description of the influence of different color parameters on the visual effect. Through this mapping table, the corresponding RGB brightness values are found under different input conditions to match the perception characteristics of the human eye to light and are used for reference and adjustment in subsequent control. After constructing the complete parameter mapping table, the parameter adjustment strategy is optimized to ensure that the adjustments between different parameters do not interfere with each other. Based on this mapping table, small perturbations within the preset error range are made, and a double-input and double-output perturbation observation mechanism is constructed. This mechanism measures the interactive influence between parameters by applying small perturbations to the lighting parameters and observing the changes in the visual effect. For example, on the basis of the currently set hue and saturation, a small perturbation amount is added to or subtracted from the lightness and brightness respectively and , then the changes in the visual effect are measured, and the data differences before and after the perturbation are recorded. These data will be used to calculate the sensitivity between parameters, thereby establishing a more accurate parameter decoupling model. After obtaining the data on the changes in the visual effect before and after the perturbation, the local Jacobian matrix between parameters is calculated based on these data. This matrix is used to describe the sensitivity relationship of the current lighting parameters, and the calculation formula of this matrix is as follows:

[0069]

[0070] where represents the partial derivative of the th visual effect parameter (such as brightness, hue) with respect to the th control parameter (such as RGB value), represents the small change amount of the visual effect, and represents the small perturbation value of the input lighting parameter. Each element of the Jacobian matrix reflects the degree of influence of a control variable on the visual effect. When the value of the matrix is large, it indicates that this parameter has a strong influence on visual perception, while when the matrix value is small, it indicates that the adjustment of this parameter has a small influence on the visual effect. Due to the coupling influence between parameters in this matrix, its inverse matrix is calculated to construct a parameter compensator, thereby realizing the decoupling of lighting parameters. The role of the compensator is to offset the cross influence between different parameters through matrix transformation methods, so as to ensure that when adjusting a certain parameter, it will not interfere with other parameters. The decoupling transformation matrix is calculated from the inverse matrix of the Jacobian matrix , that is: This matrix is used to correct the control signal so that the input lighting parameter adjustment can act independently on hue, saturation, lightness, and brightness without affecting each other. After calculating the decoupling transformation matrix, the initial lighting parameters are determined according to the driving situation feature vector and the situation correlation matrix, and the decoupling transformation matrix is used to adjust them to ensure that the finally output lighting parameters can adapt to the current driving environment. According to the driving situation category, the corresponding adjustment weights of the ambient light parameters are extracted from the situation correlation matrix, and combined with information such as lighting conditions, driver status, vehicle speed change rate, and environmental complexity in the driving situation feature vector, the set values of the initial lighting parameters are determined. These set values include the default values of the basic hue, saturation, lightness, and brightness, and are dynamically adjusted in combination with the situation information. The initial lighting parameters are input into the decoupling transformation matrix for processing to automatically calculate and adjust the mutual influence between the parameters, and the final decoupled lighting parameters are obtained. In this process, the decoupling transformation matrix is used to perform a linear transformation on the initial lighting parameters to eliminate the dependence between the parameters and ensure that the hue, saturation, lightness, and brightness are adjusted independently. The finally generated decoupled lighting parameters can provide the best visual experience in the current driving environment and are adaptively adjusted according to subsequent driving situation changes to ensure the control effect of the in-vehicle ambient light based on environmental perception.

[0071] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0072] For E typical driving situations, a situation-illumination mapping database is constructed to obtain lighting template parameters including four basic parameters of hue, saturation, lightness, and brightness and their change ranges;

[0073] Fuzzy logic control is performed according to the driving situation feature vector to obtain the initial adjustment amount of the lighting parameters, and based on the real-time monitored driving situation change rate, the initial adjustment amount of the lighting parameters is dynamically optimized to obtain smooth adjustment parameters;

[0074] The interior space of the vehicle is divided into 4 lighting areas, including the driving area, the co-pilot area, the left rear row, and the right rear row, and the smooth adjustment parameters are processed with regional differences to obtain regionalized lighting control parameters;

[0075] By comparing the Euclidean distance between the regionalized lighting control parameters and the lighting template parameters, the lighting effect matching degree is obtained;

[0076] When the lighting effect matching degree is lower than the preset target score, the regionalized lighting control parameters are finely adjusted by means of binary search until the matching degree reaches the threshold to obtain the initial lighting control parameters;

[0077] Dynamic situation weighted loss calculation is performed on the initial lighting control parameters to obtain the target lighting control parameters.

[0078] Specifically, a situation-illumination mapping database is constructed for E typical driving scenarios, and the optimal illumination parameter combinations for different driving scenarios are established, including four basic parameters, namely hue, saturation, lightness, and brightness, and their adjustable ranges, to ensure that the system dynamically adjusts the lighting scheme of the in-vehicle ambient light according to different environmental requirements. The construction of the mapping database is based on a large number of driving experiments and user feedback data. During the experiments, the in-vehicle lighting environment in typical scenarios such as normal driving, highway cruising, urban congestion, night driving, and rainy driving is collected, and at the same time, the subjective comfort evaluation and visual fatigue state of the driver are recorded. By analyzing these data, the optimal illumination parameter ranges for each scenario are summarized. Fuzzy logic control is performed based on the current driving scenario feature vector to calculate the initial adjustment amount of the illumination parameters. The input variables of the fuzzy logic control include the driving scenario category, environmental light intensity, driver state, and vehicle speed change rate, while the output variables include the hue adjustment amount, saturation adjustment amount, lightness adjustment amount, and brightness adjustment amount. Each input variable is divided into multiple fuzzy sets. For example, the vehicle speed change rate is divided into stable, fluctuating, and drastic changes, and the environmental light intensity is divided into low light, normal light, and strong light, and reasoning is carried out in combination with the preset fuzzy rules. Based on the real-time monitored driving scenario change rate, dynamic optimization processing is performed on the initial adjustment amount of the illumination parameters. When the driving scenario change rate is small, a smaller adjustment step is adopted to make the change of the ambient light smoother, so as to avoid discomfort caused to the driver by sudden lighting changes. When the driving scenario change rate is large, such as when the vehicle enters a tunnel from daytime or from highway to urban congestion state, the adjustment speed is accelerated to ensure that the ambient light can quickly adapt to the new driving scenario, thereby improving the adaptability and coherence of the lighting adjustment. To optimize the lighting control effect, the in-vehicle space is divided into four independent lighting areas, including the driving area, the co-pilot area, the left rear row, and the right rear row, and the calculated smooth adjustment parameters are processed with regional differences to meet the needs of occupants in different areas. According to the activity areas and needs of the driver and passengers, the lighting parameters of each area are adjusted independently. After calculating the regional lighting control parameters, the matching degree between the current lighting scheme and the preset optimal lighting scheme is evaluated, the Euclidean distance between the regional lighting control parameters and the lighting template parameters is calculated, and the lighting effect matching degree is evaluated according to the calculation results. The higher the matching degree, the closer the current lighting scheme is to the best setting. When the lighting effect matching degree is lower than the preset target score, the system will automatically enter the optimization mode to further fine-tune the lighting parameters.During the optimization process, a binary search method is used to fine-tune the regional lighting control parameters, calculate the adjustment direction of the current lighting parameters, and attempt to find a better parameter combination in the parameter space. Each time an adjustment is made, the new Euclidean distance is calculated, and it is judged whether the matching degree has improved. The system will continuously execute this process until the matching degree reaches the preset threshold or the number of adjustments reaches the maximum allowable value. With the support of the optimization strategy, the system can ensure that the final lighting parameters are as close as possible to the best settings to improve the adaptability and consistency of the in-vehicle lighting experience. After optimizing the lighting parameters, the dynamic scenario weighted loss is calculated to ensure that the final target lighting control parameters adapt to the specific requirements of different driving scenarios. To this end, a driving experience evaluation model is constructed, and the current lighting effect is scored according to four dimensions: visual comfort, driving assistance, mood regulation, and scenario matching. The weight of each dimension is initially set to 0.25 and dynamically adjusted based on the driver's facial expressions, eye movements, and operating behaviors captured by the in-vehicle camera. The scoring range is set between 0 and 100 to reflect the adaptability of the lighting scheme in different driving scenarios. After obtaining the driving experience score, the dynamic scenario weighted loss function is used to calculate the final target lighting control parameters. This loss function is weighted by the importance of different driving scenarios to optimize the lighting control strategy. During the calculation of the loss function, the adjustment of different lighting parameters is optimized according to the importance weight of the current driving environment. Through adaptive adjustment, it is ensured that the final lighting control parameters not only conform to the standard lighting template but also can be optimized according to the driver's real-time state and external environment changes to provide the best driving experience. Through the above steps, the target lighting control parameters are calculated and transmitted to the in-vehicle ambient light hardware execution unit, enabling it to automatically adjust the lighting scheme in different driving scenarios.

[0079] In a specific embodiment, the process of performing steps to calculate the dynamic scenario weighted loss of the initial lighting control parameters to obtain the target lighting control parameters may specifically include the following steps:

[0080] Construct an evaluation model based on the driver's expression, eye movement, and operating behavior, and set four evaluation dimensions: visual comfort, driving assistance, mood regulation, and scenario matching to obtain the experience score;

[0081] Apply the dynamic scenario weighted loss function to calculate the initial lighting control parameters to obtain the initial loss value, and perform the gradient descent optimization algorithm based on the initial loss value to obtain the optimized lighting parameters;

[0082] Conduct a safety constraint test on the optimized lighting parameters to obtain lighting parameters that meet the safety standards;

[0083] Compare the lighting parameters that meet the safety standards with the initial lighting control parameters through loss function evaluation to obtain the optimization effect data;

[0084] Based on the optimization effect data and experience scores, use the incremental learning method to update the situation-illumination mapping database and output the target lighting control parameters.

[0085] Specifically, a comprehensive evaluation model is constructed to analyze the driver's state, and based on this, the control parameters of the in-vehicle ambient light are adjusted to provide a more intelligent and personalized lighting solution. In this model, the driver's facial expressions are captured in real time through an in-vehicle camera, and deep learning algorithms are used for facial expression classification to identify the driver's current emotional state, such as nervous, relaxed, fatigued, or focused. At the same time, the system integrates eye-tracking technology to analyze key indicators such as the driver's eye movement trajectory, blink frequency, and fixation time to evaluate their attention level. The driver's operating behavior, including the steering wheel rotation angle, braking force, and throttle control method, is monitored through the in-vehicle control unit to further judge the driver's driving style and state. After these information are fused, the driving experience is divided into four core evaluation dimensions, namely visual comfort, driving assistance, emotional regulation, and situation matching. Among them, visual comfort is used to measure whether the current lighting environment interferes with the driver's visual perception, driving assistance is used to evaluate whether the lighting conditions help improve driving safety, emotional regulation reflects the impact of the lighting solution on the driver's emotional state, and situation matching measures whether the current lighting solution is suitable for the driving situation. The scores of these four dimensions together constitute the overall experience score of the driver, and the score range is set between 0 and 100 to quantitatively evaluate the adaptability of different lighting parameters. After obtaining the experience score, a dynamic situation weighted loss function is applied to calculate the initial lighting control parameters to quantify the deviation between the lighting parameters and the best experience. The loss function optimizes the adjustment of the lighting parameters according to the importance of different driving situations. Calculate the difference between the current lighting solution and the best lighting solution and map it into the loss function, where the loss value is calculated as follows: The system performs a gradient descent optimization algorithm based on this loss value, calculates the partial derivative of the loss function with respect to each lighting parameter in each iteration, and adjusts the parameter values according to the gradient direction to make it continuously approach the optimal solution. In the specific implementation process, the system sets the learning rate to 0.01, the maximum number of iterations to 100, and the convergence threshold to 0.001 to ensure that the optimization process converges to the optimal solution at a reasonable computational cost. After this optimization process, the optimized lighting parameters are generated. Perform a safety constraint test on the optimized lighting parameters to ensure that the final lighting control scheme meets the road safety standards and the driver's visual comfort requirements. The safety constraint test mainly includes brightness limit, color temperature regulation, and hue change rate constraint. Among them, the brightness limit ensures that the light intensity of the in-vehicle ambient light does not exceed the maximum brightness specified by the regulations to prevent affecting the driver's or other road users' line of sight, and the color temperature regulation is used to ensure that the lighting solution conforms to the human eye's visual comfort range and avoid visual fatigue or glare problems caused by too high or too low color temperature.In addition, a hue change rate constraint is set to ensure that the color change of the ambient light does not exceed the set threshold within the specified time, thereby avoiding the impact of sudden lighting on the driver's attention. After this series of safety inspections, lighting parameters that meet safety standards are obtained. The lighting parameters that meet safety standards are compared with the initial lighting control parameters through loss function evaluation to quantify the actual effect of the optimization process. Calculate the difference in loss values before and after optimization to evaluate the improvement amplitude of the optimization strategy. If the loss value after optimization is significantly reduced, it indicates that the optimization strategy effectively improves the adaptability of the lighting scheme. If the change in the loss value is not significant, it indicates that the current optimization strategy needs further adjustment. In addition, calculate the difference in experience scores before and after optimization to verify whether the optimization improves the driver's actual experience, and evaluate the impact of different lighting parameters on the driving experience through statistical analysis. All these evaluation data will be stored as optimization effect data for subsequent system learning and optimization. To ensure that the system can continuously learn and evolve to adapt to changes in different driving environments, based on the optimization effect data and experience scores, an incremental learning method is used to update the scenario-lighting mapping database. This update process adopts a batch update method, and the database is adjusted once every 100 driving data are accumulated. During the database update process, combined with the newly collected data, recalculate the optimal lighting parameters in different scenarios, and perform regression analysis based on historical optimization data to ensure that the updated database will not be over-adjusted due to short-term data deviation. During the incremental learning process, an adaptive weight adjustment strategy is used to assign higher weights to long-term stable data and lower weights to data with large recent changes to ensure the stability and adaptability of database updates. The system outputs the optimized target lighting control parameters and transmits them to the in-vehicle ambient light hardware execution unit through the CAN bus to achieve precise lighting control.

[0086] The above describes the vehicle-mounted ambient light control method based on environment perception in the embodiments of the present invention. Next, the vehicle-mounted ambient light control system based on environment perception in the embodiments of the present invention will be described. Please refer to Figure 2 , an embodiment of the vehicle-mounted ambient light control system based on environment perception in the embodiments of the present invention includes:

[0087] An acquisition module 201, configured to acquire in-vehicle image data and vehicle-mounted sensor data and input them into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fused feature map;

[0088] A scenario recognition module 202, configured to perform scenario recognition and analysis on the fused feature map to obtain a driving scenario feature vector and a scenario association matrix;

[0089] The decoupling processing module 203 is configured to perform decoupling processing on the in-vehicle atmosphere lamp parameters according to the driving scenario feature vector and the scenario correlation matrix to obtain the decoupled lighting parameters;

[0090] The loss calculation module 204 is configured to perform scenario adaptive lighting regulation and dynamic scenario weighted loss calculation on the decoupled lighting parameters and the driving scenario feature vector to obtain the target lighting control parameters.

[0091] Through the collaborative cooperation of the above-mentioned various components, by adopting the dual-branch Mamba information fusion module to process the in-vehicle image data and the in-vehicle sensor data, the effective fusion of multi-source heterogeneous data is achieved, and the perception ability of the driving scenario is improved. This module uses the image processing branch and the sensor data processing branch to extract spatial features and temporal features respectively, and performs adaptive fusion through the cross-modal feature fusion layer, overcoming the problem of insufficient information of a single data source. By adopting the multi-scale driving scenario attention module, multi-scale feature representations are extracted through convolutional kernels of different sizes, and the self-attention mechanism and cross-scale attention fusion are applied to achieve accurate recognition and analysis of different driving scenarios, enabling the system to focus on the features most relevant to the atmosphere lamp adjustment. Aiming at the color-brightness coupling problem existing in the traditional in-vehicle atmosphere lamp controller, the present invention proposes a parameter decoupling processing mechanism. By establishing an LAFC parameter coupling relationship model and a dual-input dual-output disturbance observation mechanism, independent and accurate control of the four parameters of hue, saturation, lightness, and brightness is achieved, making the adjustment of the in-vehicle atmosphere lamp more accurate. Based on the scenario adaptive lighting regulation method, the system can dynamically adjust the lighting parameters according to the identified driving scenario and perform differential control on different areas in the vehicle, meeting the personalized lighting needs of different occupants in different driving scenarios. By innovatively introducing a dynamic scenario weighted loss function, the problem of unbalanced lighting requirements in different driving scenarios is solved. By adjusting the scenario weight factor, the system can perform priority adjustment for specific driving scenarios and optimize the overall lighting effect. In addition, the system also continuously accumulates driving experience through the lighting strategy self-learning module and continuously improves the lighting control effect.

[0092] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is made to execute the steps of the in-vehicle atmosphere lamp control method based on environment perception.

[0093] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, system, and unit can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0094] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to enable an in-vehicle ambient light control device based on environmental perception (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0095] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A vehicle ambient light control method based on environment perception, characterized in that: include: Collect in-vehicle image data and on-board sensor data and input them into the dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map; The fused feature map is subjected to situation identification and analysis to obtain a driving situation feature vector and a situation association matrix, including: extracting features from the fused feature map through three convolution kernels of different sizes to obtain feature representations of three different scales; applying a self-attention mechanism to the feature representations of the three different scales, using a scaled dot product attention algorithm, calculating the spatial correlation within the feature map, and obtaining attention weighted feature maps of three different scales; inputting the attention weighted feature maps of the three different scales into a cross-scale attention fusion module, and adaptively fusing them through learnable weight parameters to obtain a fused feature vector; performing full connection processing on the fused feature vector to output a driving situation category distribution; based on the driving situation category distribution, extracting target feature indicators of each driving situation to obtain a driving situation feature vector, wherein the target feature indicators include lighting conditions, driver status, vehicle speed change rate, and environmental complexity; performing quantitative analysis on the driving situation feature vector, calculating the correlation coefficients between E types of driving situations and four types of ambient light parameters, and generating a situation association matrix, wherein the four types of ambient light parameters include hue, saturation, brightness, and brightness; Decoupling the vehicle ambient light parameters according to the driving scenario feature vector and the scenario association matrix to obtain decoupled lighting parameters; Context-adaptive lighting regulation and dynamic context-weighted loss calculation are performed on the decoupled lighting parameters and the driving context feature vector to obtain target lighting control parameters.

2. The vehicle-mounted ambient light control method based on environment perception according to claim 1, characterized in that: The in-vehicle image data and the vehicle-mounted sensor data are collected and input into the dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map, including: The image of the interior environment of the vehicle is collected by a high-resolution camera installed at the rearview mirror inside the vehicle to obtain an in-vehicle camera image; The vehicle-mounted sensor array is used to collect environmental parameters, wherein the vehicle-mounted sensor array includes a light sensor, a temperature sensor, an acceleration sensor and a GPS positioning module to obtain the vehicle surrounding environmental parameter data; Packaging the in-vehicle camera image and the vehicle surrounding environment parameter data according to a predefined data packet format to obtain a formatted data packet, wherein the predefined data packet format includes a timestamp, a sensor ID, a data value, and a status identifier; Dividing the formatted data packets into a real-time high-priority data stream and a scheduled low-priority data stream to obtain a hierarchical data stream, and dynamically adjusting the hierarchical data stream according to the driving state of the vehicle to obtain in-vehicle image data and vehicle-mounted sensor data; The in-vehicle image data and the on-board sensor data are input into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map.

3. The vehicle-mounted ambient light control method based on environment perception according to claim 2 is characterized in that: The in-vehicle image data and the vehicle-mounted sensor data are input into a dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map, including: Performing image preprocessing on the in-vehicle image data to obtain a standardized image, and performing data preprocessing on the in-vehicle sensor data to obtain standardized sensor data; Input the standardized image into the image processing branch of the dual-branch Mamba information fusion module, extract spatial features through 4 convolutional layers, and obtain image spatial features; Input the standardized sensor data into the sensor data processing branch of the dual-branch Mamba information fusion module, extract the timing features through the Mamba state space model, and set the state dimension to E8, the expansion step to 4, the expansion coefficient to 2, and the bidirectional scanning mode to obtain the sensor timing features; Input the image spatial features and the sensor temporal features into the cross-modal feature fusion layer of the dual-branch Mamba information fusion module, calculate the attention weight matrix using the attention mechanism and normalize it through the Softmax function, and obtain the adaptive fusion features; A residual connection is added to the adaptive fusion feature to retain the original feature information, and layer normalization is used to prevent feature distribution deviation to obtain a fusion feature map.

4. The vehicle-mounted ambient light control method based on environment perception according to claim 1, characterized in that: The decoupling process of the vehicle ambient light parameters according to the driving scenario feature vector and the scenario association matrix to obtain the decoupled lighting parameters includes: By experimentally measuring the visual effects under different RGB color combinations and brightness levels, a LAFC parameter coupling relationship model is established to obtain a coupling coefficient matrix, which represents the degree of mutual influence between the four parameters of hue, saturation, lightness and brightness; The HSV color space is divided into a plurality of grids based on the coupling coefficient matrix, and the corresponding relationship between the visual effect and the RGB brightness is calculated in each grid to obtain a four-dimensional parameter mapping table; Based on the four-dimensional parameter mapping table, a small disturbance within a preset error range is performed to construct a dual-input dual-output disturbance observation mechanism to obtain visual effect change data before and after the disturbance; According to the visual effect change data before and after the disturbance, the local Jacobian matrix between the parameters is calculated to obtain a parameter sensitivity matrix, and a parameter compensator is constructed based on the inverse matrix of the parameter sensitivity matrix to obtain a decoupling transformation matrix; Initial lighting parameters are determined according to the driving scenario feature vector and the scenario association matrix, and the initial lighting parameters are input into the decoupling transformation matrix for processing, and the decoupled lighting parameters are obtained by automatically calculating the mutual influence between the adjustment parameters.

5. The vehicle-mounted ambient light control method based on environment perception according to claim 1, characterized in that: The performing situation-adaptive lighting control and dynamic situation weighted loss calculation on the decoupled lighting parameters and the driving situation feature vector to obtain target lighting control parameters includes: For E typical driving scenarios, a scenario-lighting mapping database is constructed to obtain lighting template parameters including four basic parameters of hue, saturation, brightness, and brightness and their range of variation; Executing fuzzy logic control according to the driving situation feature vector to obtain an initial adjustment amount of the lighting parameter, and dynamically optimizing the initial adjustment amount of the lighting parameter based on a driving situation change rate monitored in real time to obtain a smooth adjustment parameter; The interior space of the vehicle is divided into four lighting areas, including a driving area, a co-pilot area, a rear left side, and a rear right side, and the smooth adjustment parameters are processed by regional differentiation to obtain regional lighting control parameters; Obtaining a lighting effect matching degree by comparing the Euclidean distance between the regionalized lighting control parameter and the lighting template parameter; When the lighting effect matching degree is lower than the preset target score, the regionalized lighting control parameters are fine-tuned by binary search until the matching degree reaches a threshold value, thereby obtaining initial lighting control parameters; Dynamic situational weighted loss calculation is performed on the initial lighting control parameters to obtain target lighting control parameters.

6. The vehicle-mounted ambient light control method based on environment perception according to claim 5, characterized in that: The performing dynamic situation weighted loss calculation on the initial lighting control parameter to obtain the target lighting control parameter includes: An evaluation model is built based on the driver's facial expressions, eye movements and operating behaviors, and four evaluation dimensions are set: visual comfort, driving assistance, emotional regulation, and situation matching, to obtain an experience score. Applying a dynamic context weighted loss function to calculate the initial lighting control parameters to obtain an initial loss value, and executing a gradient descent optimization algorithm based on the initial loss value to obtain optimized lighting parameters; Performing a safety constraint check on the optimized lighting parameters to obtain lighting parameters that meet safety standards; Performing loss function evaluation and comparison on the lighting parameters that meet the safety standards and the initial lighting control parameters to obtain optimization effect data; Based on the optimization effect data and the experience score, the scenario-lighting mapping database is updated using an incremental learning method, and target lighting control parameters are output.

7. A vehicle-mounted ambient light control system based on environmental perception, characterized in that: Used to implement the vehicle-mounted ambient light control method based on environment perception according to any one of claims 1 to 6, the vehicle-mounted ambient light control system based on environment perception comprises: The acquisition module is used to collect in-vehicle image data and vehicle sensor data and input them into the dual-branch Mamba information fusion module for feature extraction and fusion processing to obtain a fusion feature map; A situation recognition module is used to perform situation recognition and analysis on the fused feature map to obtain a driving situation feature vector and a situation association matrix, including: extracting features from the fused feature map through three convolution kernels of different sizes to obtain feature representations of three different scales; applying a self-attention mechanism to the feature representations of the three different scales, using a scaled dot product attention algorithm, calculating the spatial correlation within the feature map, and obtaining attention weighted feature maps of three different scales; inputting the attention weighted feature maps of the three different scales into a cross-scale attention fusion module, and adaptively fusing them through learnable weight parameters to obtain a fused feature vector; performing full connection processing on the fused feature vector to output a driving situation category distribution; based on the driving situation category distribution, extracting target feature indicators for each driving situation to obtain a driving situation feature vector, wherein the target feature indicators include lighting conditions, driver status, vehicle speed change rate, and environmental complexity; performing quantitative analysis on the driving situation feature vector, calculating the correlation coefficients between E types of driving situations and four types of ambient light parameters, and generating a situation association matrix, wherein the four types of ambient light parameters include hue, saturation, brightness, and brightness; A decoupling processing module, used for performing decoupling processing on vehicle ambient light parameters according to the driving scenario feature vector and the scenario association matrix to obtain decoupled lighting parameters; The loss calculation module is used to perform situation-adaptive lighting regulation and dynamic situation weighted loss calculation on the decoupled lighting parameters and the driving situation feature vector to obtain target lighting control parameters.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the processor is enabled to execute the vehicle ambient light control method based on environment perception as claimed in any one of claims 1 to 6.

Images