Ambient illuminance estimation method and system and storage medium

Abstract

The invention provides an environment illuminance estimation method and system and a storage medium. The method comprises the steps that image data of the environment where a vehicle is located are collected in real time; selecting a plurality of ROI regions from the image data according to a preset ROI selection algorithm; acquiring environment feature data based on the ROI region; and estimating final environment illuminance data through a preset illuminance estimation model based on the environment characteristic data. According to the method provided by the invention, the technical problems of high cost, limited installation position, incapability of providing scene information and single function of a traditional light sensor for detecting the ambient illuminance are solved. According to the invention, the estimation cost of the ambient illuminance is effectively reduced, the accuracy and reliability of the detection of the ambient illuminance are improved, and rich scene information and multifunctionality are provided for vehicle-mounted applications.

Description

Technical Field

[0001] This application belongs to the field of automotive electronics technology, and particularly relates to an environmental illuminance estimation method, system, and storage medium. Background Art

[0002] Currently, most automobiles are equipped with light sensors to detect environmental illuminance and control the automatic turning on and off of vehicle lights. These traditional light sensors usually have many limitations, such as high cost, limited installation location, inability to provide scene information, single function, etc. Therefore, there is an urgent need for a more intelligent, reliable, cost-effective, and efficient illuminance estimation method based on camera images to replace traditional automotive light sensors and provide more abundant functions. Summary of the Invention

[0003] To solve the above technical problems, this application proposes an environmental illuminance estimation method, system, and storage medium with low cost, diverse functions, high detection accuracy, and strong reliability.

[0004] Specifically, this application proposes an environmental illuminance estimation method, including:

[0005] Real-time collecting image data of the environment where the vehicle is located; selecting multiple ROI (Region of Interest) regions from the image data according to a preset ROI selection algorithm; obtaining environmental feature data based on the ROI regions.

[0006] And, estimating the final environmental illuminance data based on the environmental feature data through a preset illuminance estimation model.

[0007] In the above technical solution, by selecting multiple ROI regions through a preset ROI selection algorithm and combining with an illuminance estimation model for estimation, the environmental illuminance can be estimated more accurately, avoiding the deviation of single-region illuminance estimation, comprehensively considering the illumination characteristics of multiple regions, and improving the illuminance estimation accuracy. By using the ROI selection algorithm to select key regions, the algorithm can adapt to changing environmental illumination conditions and still accurately estimate the illuminance in complex environments. After obtaining environmental feature data based on the ROI regions and then obtaining the final illuminance data through the illuminance estimation model, the obtained illuminance data can be more accurate and reliable. The hardware cost is effectively reduced, and by using the existing camera equipment of the vehicle, the high cost of using traditional light sensors is avoided. By combining image data and an illuminance estimation model, multi-functional environmental perception is realized.

[0008] As an implementation, after the real-time collection of image data of the environment where the vehicle is located, it further includes:

[0009] Perform distortion correction processing and format adjustment on the real-time acquired image data to obtain the final image data.

[0010] By performing distortion correction processing on the acquired image data, the distortion in the image can be eliminated, and the image quality can be improved. By adjusting the format to a unified image format and resolution, the consistency of the image is enhanced. The image data after distortion correction processing and format adjustment can improve the accuracy of subsequent processing, and enhance the robustness and precision of subsequent algorithms. By performing distortion correction processing and format adjustment, the algorithm can adapt to the image data acquired by different cameras.

[0011] Further, the selecting multiple ROI regions from the image data according to the preset ROI selection algorithm includes:

[0012] Obtain the current time information and the current vehicle state information in real time; based on the current time information and the current vehicle state information, use the preset ROI selection algorithm to determine multiple coordinate ranges according to the final image data as the corresponding ROI regions.

[0013] By using the preset ROI selection algorithm in combination with the current time information and the vehicle state information, the interference of irrelevant regions is avoided, and the accuracy of ROI region selection is improved. Combining the time information and the vehicle state information enables the algorithm to adapt to the ROI selection requirements in different environments. Optimizing the selection of ROI regions according to the vehicle state information reduces unnecessary consumption of computing resources and improves the resource utilization efficiency.

[0014] Further, the obtaining the environmental feature data based on the ROI regions includes:

[0015] Obtain the coordinate information and the region size information of the corresponding ROI region according to the coordinate range, and extract the corresponding pixel data from the image data based on the coordinate information and the region size information.

[0016] Convert the pixel data into grayscale image data; generate a brightness distribution histogram based on the grayscale image data; obtain the environmental feature data of the ROI region based on the brightness distribution histogram.

[0017] By extracting the pixel data of the ROI region to generate a brightness distribution histogram, the illumination characteristics of this region can be accurately reflected, facilitating subsequent brightness analysis and processing. By converting the pixel data into grayscale image data, the data volume is reduced, and the data processing efficiency is improved. Through the brightness distribution histogram, rich environmental illumination information can be obtained, providing effective data support for environmental perception. The environmental feature data obtained through the brightness distribution histogram can improve the accuracy and reliability of illuminance estimation.

[0018] Further, the illuminance estimation model includes at least one of a machine learning model or a mapping function model; estimating the environmental illuminance data through the preset illuminance estimation model includes:

[0019] Based on the current time information, the current vehicle state information, and the environmental feature data, use the pre-trained illuminance estimation model to estimate the first environmental illuminance data.

[0020] Based on the first environmental illuminance data, adjust the estimation parameters of the illuminance estimation model in real time.

[0021] Through the machine learning model or the mapping function model combined with the current time information, vehicle state information, and environmental feature information, the environmental illuminance data can be estimated more accurately. Adjusting the model parameters based on the first environmental illuminance data enables the illuminance estimation model to adapt to the illuminance estimation requirements in different environments, can dynamically adjust the model parameters, and improves the robustness and estimation accuracy of the illuminance estimation model.

[0022] Further, after estimating the first environmental illuminance data, it further includes:

[0023] Identify the glare area from the first environmental illuminance data based on a preset illuminance threshold.

[0024] Perform a weight reduction process or a shielding process on the first environmental illuminance data in the glare area to obtain the second environmental illuminance data.

[0025] Performing anti-glare processing on the first environmental illuminance data provides a multi-functional effect for illuminance estimation. Identifying glare through a preset illuminance threshold and performing a weight reduction process or a shielding process on the identified glare can effectively prevent the influence of glare on the illuminance estimation result, thereby improving the reliability and accuracy of the final illuminance data. By performing a weight reduction process or a shielding process on the glare area, the unstable factors in the illuminance estimation process can be effectively reduced, and more stable and reliable environmental illuminance data can be provided. Unnecessary data processing is reduced, and the efficiency of environmental

[0026] illuminance data estimation is effectively improved.

[0027] Further, after obtaining the second environmental illuminance data, it further includes:

[0028] Obtain the current weather information in real time; perform corresponding adaptation adjustment processing on the second environmental illuminance data based on the current weather information to obtain the final environmental illuminance data.

[0029] Among them, the adaptation adjustment processing includes at least denoising, defogging, and / or multi-frame fusion processing.

[0030] By adaptively adjusting the second ambient light intensity data in combination with the current weather information, the estimation error of the light intensity caused by weather factors can be eliminated, so that the obtained final ambient light intensity data can more truly reflect the current ambient light intensity condition, improving the accuracy and reliability of the estimation of the final ambient light intensity data, enabling the algorithm to adapt to different weather changes, ensuring the effectiveness and stability of the ambient light intensity data, and avoiding judgment errors or instability caused by weather changes. Through multi-frame fusion processing, the errors or noises that may exist in single-frame data are overcome, making the light intensity estimation result more accurate.

[0031] Further, after obtaining the final ambient light intensity data, the following steps are also included:

[0032] Convert the final ambient light intensity data into a preset standard format and output the converted final ambient light intensity data to the in-vehicle application program.

[0033] By converting the final ambient light intensity data into a preset standard format, the compatibility between the final ambient light intensity data and the in-vehicle application program is ensured, avoiding the problem of data format mismatch. The integration of the algorithm and the in-vehicle application program is enhanced. The ambient light intensity data in the standard format can be efficiently utilized by the in-vehicle application program, improving the data utilization efficiency. The ambient light intensity data in the standard format can support the requirements of multiple in-vehicle application programs, ensuring the effectiveness of the final ambient light intensity data.

[0034] Based on the same inventive concept, the present application also proposes a system for an ambient light intensity estimation method, and the system includes:

[0035] A data acquisition module for real-time acquisition of image data of the environment where the vehicle is located.

[0036] A region selection module for selecting multiple ROI regions from the image data according to a preset ROI selection algorithm.

[0037] A feature acquisition module for obtaining ambient feature data based on the ROI regions.

[0038] And an estimation module for estimating the final ambient light intensity data based on the ambient feature data through a preset light intensity estimation model.

[0039] Based on the same inventive concept, the present application also proposes a computer-readable storage medium, and the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions can be read and executed by a control processor to perform the ambient light intensity estimation method.

[0040] Compared with the prior art, the present application has at least the following beneficial effects:

[0041] The method proposed in this application effectively solves the technical problems of high cost, limited installation location, and single function existing in traditional optical sensors. By presetting the ROI selection algorithm to select multiple ROI regions, and combining with the illuminance estimation model for estimation, it can more accurately estimate the ambient illuminance, avoid the deviation of single-region illuminance estimation, comprehensively consider the illumination characteristics of multiple regions, and improve the accuracy of illuminance estimation. By using the ROI selection algorithm to select key regions, the algorithm can adapt to changing ambient lighting conditions and still accurately estimate the illuminance in complex environments. After obtaining the environmental feature data based on the ROI region and then obtaining the final illuminance data through the illuminance estimation model, the obtained illuminance data can be more accurate and reliable. It effectively reduces the hardware cost. By using the existing camera equipment of the vehicle, it avoids the high cost of using traditional optical sensors. By combining image data and the illuminance estimation model, it realizes multi-functional environmental perception. Brief Description of the Drawings

[0042] Figure 1 It is a flowchart of the ambient illuminance estimation method shown in the embodiments of this application.

[0043] Figure 2 It is a schematic diagram of the ambient illuminance estimation system shown in the embodiments of this application. Detailed Embodiments

[0044] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0045] It should be noted that the terms "first", "second", etc. in the description and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order different from those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server including a series of steps or units does not necessarily have to be limited to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0046] Embodiment 1:

[0047] Please refer toFigure 1 , the method for estimating the ambient light intensity mainly includes steps S1 to S4.

[0048] Among them, step S1 includes: real-time collecting image data of the environment where the vehicle is located. Among them, the image data can mainly be collected in real time by an in-vehicle camera. The image data can mainly be the image data of the scene in front of the vehicle. After the image data is collected, preprocessing needs to be performed on the image data, such as performing distortion correction processing on the image data, adjusting the size of the image data, and adjusting the image data to the RGB format or other formats.

[0049] Step S2 includes: selecting multiple ROI regions from the image data according to a preset ROI selection algorithm. Among them, the preset ROI selection algorithm can mainly be an ROI selection algorithm based on image features, such as an edge detection algorithm or a color segmentation algorithm. It can also be an ROI selection algorithm based on object detection, such as YOLO (You Only Look Once), Faster R-CNN (Faster Region-based Convolutional Neural Network). It can also be an ROI selection algorithm based on machine learning, such as a classifier and a clustering algorithm. Those skilled in the art can select other ROI selection algorithms according to the actual situation and are not limited to this. When presetting the ROI selection algorithm, it can also be preset according to time information, vehicle status information, and scene information. For example, when presetting the ROI selection algorithm based on time information, the road surface area in the image data is selected as the ROI region during the day, and the sky or both sides of the road are selected as the ROI region at night.

[0050] Step S3 includes: obtaining ambient feature data based on the ROI regions. Among them, it mainly includes extracting pixel data from the image data of the ROI regions, converting the pixel data into a grayscale image, generating a luminance distribution histogram based on the grayscale image, and extracting feature data such as the mean, peak, and median of the grayscale image data according to the luminance distribution histogram. The ambient feature data can at least include feature data such as the mean, peak, and median of the grayscale image data.

[0051] And, step S4: Estimate the final ambient illuminance data based on the environmental feature data through a preset illuminance estimation model. The illuminance estimation model can mainly be a machine learning model or a mapping function model. By using a pre-trained machine learning model or a predefined mapping function model, the final ambient illuminance data is estimated according to the environmental feature data. After obtaining the final ambient illuminance data, the illuminance estimation model is optimized based on the final ambient illuminance data to improve the accuracy and reliability of the estimation of the ambient illuminance data.

[0052] That is, the image data of the scene in front of the vehicle can be collected in real time by the in-vehicle front camera. After preprocessing the image data, multiple ROI regions can be selected from the processed image data through a ROI selection algorithm based on machine learning; Pixel data is extracted from the corresponding image data based on the ROI region, the pixel data is converted into a grayscale image, and a luminance distribution histogram is generated based on the grayscale image. Feature data such as mean, median, and peak are obtained according to the luminance distribution histogram. Illuminance estimation is performed using a pre-trained machine learning model based on the feature data to obtain the final ambient illuminance data.

[0053] In some embodiments, after collecting the image data of the environment where the vehicle is located in real time, it further includes:

[0054] Perform distortion correction processing and format adjustment on the image data collected in real time to obtain the final image data.

[0055] Performing distortion correction processing on the image data is mainly to eliminate image distortion caused by the characteristics of the camera lens or the installation angle, so as to ensure that the image can accurately reflect the actual scene. The distortion correction processing can at least include lens distortion correction, geometric correction, and dynamic distortion correction. The dynamic distortion correction can mainly be based on vehicle states, such as vehicle speed, steering angle and other information to adjust the camera distortion parameters. The format adjustment can at least include adjusting the size of the image data, the format of the image, etc., such as adjusting the image data to the RGB format or other color formats. To ensure the validity of the image data.

[0056] Optionally, the selecting multiple ROI regions from the image data according to a preset ROI selection algorithm includes:

[0057] Obtain the current time information and the current vehicle state information in real time; Based on the current time information and the current vehicle state information, use the preset ROI selection algorithm to determine multiple coordinate ranges according to the final image data as the corresponding ROI regions.

[0058] Among them, the current time information can be mainly obtained in real time through the vehicle-mounted system, and the current vehicle state information can be obtained through vehicle-mounted sensors. The preset ROI selection algorithm can mainly be based on preset selection algorithms such as time information, vehicle state information, etc. For example, when the time information is daytime, the ROI area can be set as the area of the road surface in the image; when the time information is night, the ROI area is set as the sky or both sides of the road. The ROI area can be adjusted in real time according to the vehicle state information. For example, the image area of the road surface at a long distance or the front vehicle can be selected as the ROI area according to the vehicle state information. The area with uniform color or brightness in the image data can be used as the ROI area. Those skilled in the art can choose other ways to set the ROI area according to the actual situation.

[0059] Optionally, obtaining environmental feature data based on the ROI area includes:

[0060] Obtain the coordinate information and area size information of the corresponding ROI area according to the coordinate range, and extract the corresponding pixel data from the image data based on the coordinate information and area size information.

[0061] Convert the pixel data into grayscale image data; generate a brightness distribution histogram based on the grayscale image data.

[0062] Obtain the environmental feature data of the ROI area based on the brightness distribution histogram.

[0063] Among them, the coordinate information and area size information of the ROI area are obtained according to the coordinate range of the ROI area. For example, the coordinate range of the ROI area can be (x1, y1, x2, y2), where (x1, y1) is the lower left corner coordinate of the ROI area range, and (x2, y2) is the upper right corner coordinate of the ROI area. The environmental feature data can mainly include the median, mean, peak, etc. of the histogram data, but is not limited thereto.

[0064] Optionally, the illuminance estimation model includes at least one of a machine learning model or a mapping function model; estimating the environmental illuminance data through the preset illuminance estimation model includes:

[0065] Based on the current time information, the current vehicle state information and the environmental feature data, estimate the first environmental illuminance data by using a pre-trained illuminance estimation model. Adjust the estimation parameters of the illuminance estimation model in real time based on the first environmental illuminance data.

[0066] Among them, the illuminance estimation model can be a machine learning model, such as a linear regression model, a support vector machine model, a random forest model, and a neural network model, etc. The machine learning model is trained in advance using a large amount of historical time information, historical vehicle state information, and environmental feature data, and the trained machine learning model is used to estimate the environmental illuminance data. The illuminance estimation model is adjusted in real time according to the obtained first environmental illuminance data to optimize the illuminance estimation model, so that the estimated environmental illuminance data is more accurate.

[0067] The illuminance estimation model can also be a mapping function model. For example, a linear function model or a non-linear function model can be adopted. The environmental feature data is mapped into environmental illuminance data according to a predefined mapping relationship. After obtaining the environmental illuminance data, the mapping parameters of the mapping function model are adjusted according to the environmental illuminance data to optimize the mapping function model and ensure the accuracy of the final environmental illuminance data.

[0068] Optionally, after estimating the first environmental illuminance data, it further includes:

[0069] Identifying a glare area from the first environmental illuminance data based on a preset illuminance threshold.

[0070] Performing a weight reduction process or a shielding process on the first environmental illuminance data of the glare area to obtain second environmental illuminance data.

[0071] Among them, the preset illuminance threshold can be set according to experimental data or empirical values. For example, the illuminance threshold can be set to 5000 lux, and the area where the illuminance exceeds 5000 lux is determined as the glare area. The weight reduction process can mainly use the weighted average method and the exponential decay method to reduce the weight of the glare area. It is also possible to perform a shielding process on the glare area, set the illuminance of the glare area to an invalid value, or replace the illuminance value of the glare area with the illuminance value of other non-glare areas. The illuminance data after performing the weight reduction process or the shielding process on the glare area is used as the second environmental illuminance data.

[0072] Optionally, after obtaining the second environmental illuminance data, it further includes:

[0073] Real-time obtaining the current weather information. Based on the current weather information, corresponding adaptive adjustment processing is performed on the second environmental illuminance data to obtain the final environmental illuminance data. Among them, the adaptive adjustment processing at least includes denoising, defogging, and / or multi-frame fusion processing.

[0074] Among them, the current weather information can mainly be obtained by other vehicle-mounted sensors. For example, the current weather information can be obtained through a wiper sensor, a temperature sensor, a humidity sensor, etc. When it is detected that the current weather information is rain, snow or fog, the second ambient light intensity data is processed by denoising, defogging, multi-frame fusion, etc., to avoid the estimation error of the light intensity data caused by the weather information. Effectively improve the adaptability of the light intensity estimation model to different weathers. When the current weather is rain, Gaussian filtering can be mainly used to remove raindrop noise, and then combined with multi-frame fusion processing to reduce the influence of raindrops on the light intensity data. When the current weather is fog, the contrast of the light intensity data can be enhanced mainly by the dark channel prior defogging algorithm, combined with multi-frame fusion processing to improve the accuracy of the ambient light intensity data. When the current weather is snow, median filtering can be used to remove snowflake noise, combined with multi-frame fusion processing to improve the stability of the ambient light intensity data. Those skilled in the art can select other algorithms for denoising and defogging operations according to the actual situation, and are not limited thereto.

[0075] Optionally, after obtaining the final ambient light intensity data, it further includes:

[0076] Convert the final ambient light intensity data into a preset standard format, and output the final ambient light intensity data after the conversion format to the vehicle-mounted application program.

[0077] Among them, the preset standard format can mainly be to convert the ambient light intensity data into lux units. Those skilled in the art can set the preset standard format to other formats according to the actual situation, and are not limited thereto. The vehicle-mounted application program can mainly include a vehicle lamp control system, an assisted driving system, an instrument panel brightness adjustment application program, etc., and is not limited thereto.

[0078] Embodiment 2:

[0079] Please refer to Figure 2 , this application also proposes a system using the ambient light intensity estimation method described in Embodiment 1. The system includes: a data acquisition module, a region selection module, a feature acquisition module, and an estimation module.

[0080] Among them, the data acquisition module is used to collect image data of the environment where the vehicle is located in real time. Among them, the image data of the scene in front of the vehicle can be mainly captured in real time by a vehicle-mounted camera, and each frame of image data is preprocessed. For example, the size of each frame of image data is adjusted, and the image data is subjected to distortion correction, etc., to improve the quality of the image data and ensure the accuracy of subsequent light intensity estimation.

[0081] A region selection module, configured to select multiple ROI regions from the image data according to a preset ROI selection algorithm. Among them, the preset ROI selection algorithm can be mainly set according to the current time information, vehicle status information, and current scene information. For example, setting the ROI selection algorithm according to the time information, setting the road part area in the image data as the ROI region during the day, and selecting the sky or the areas on both sides of the road as the ROI region at night. Adjusting the ROI region according to the vehicle status information. For example, when the vehicle speed is above 60 km / h, taking the road part area beyond 100 meters in front of the vehicle as the ROI region. When setting the ROI region according to the scene information, an area with uniform color or brightness can be selected as the ROI region according to the content of the image data.

[0082] A feature acquisition module, configured to acquire environmental feature data based on the ROI region. Among them, the pixel data can be mainly extracted from the image data of the ROI region, the pixel data is converted into a grayscale image, and by statistically analyzing the grayscale image, a brightness distribution histogram is generated. The environmental feature data is extracted from the brightness distribution histogram, and the environmental feature data can at least include the median, mean, peak value, etc. of the pixel data.

[0083] And an estimation module, configured to estimate the final environmental illuminance data based on the environmental feature data through a preset illuminance estimation model. Among them, the illuminance estimation model can at least adopt a machine learning model and a mapping function model. Estimating based on the environmental feature data through a pre-trained machine learning model to obtain the final environmental illuminance data. The environmental feature data can also be mapped to the environmental illuminance data by using a predefined mapping function model based on the environmental feature data.

[0084] The environmental illuminance estimation system may further include a glare processing module and a weather adaptation module.

[0085] Among them, the glare processing module is configured to identify a glare region from the first environmental illuminance data based on a preset illuminance threshold; perform a weight reduction process or a shielding process on the first environmental illuminance data of the glare region to obtain second environmental illuminance data. The weight reduction process can at least use the weighted average method to reduce the weight of the environmental illuminance data in the glare region. Those skilled in the art can select other methods to perform the weight reduction process on the first environmental illuminance data according to the actual situation, which is not limited thereto. The shielding process can at least be setting the illuminance data in the glare region as invalid data.

[0086] The weather adaptation module is used to obtain the current weather information in real time. Based on the current weather information, corresponding adaptation adjustment processing is performed on the second ambient illumination data to obtain the final ambient illumination data. Among them, the adaptation adjustment processing at least includes denoising, defogging, and / or multi-frame fusion processing. To ensure that the ambient illumination estimation model can adapt to various different weather scenarios, so as to ensure the accuracy and reliability of the estimation results.

[0087] Embodiment 3:

[0088] The present application also proposes a computer-readable storage medium, which includes:

[0089] The computer-readable storage medium stores computer-executable instructions.

[0090] When the computer-executable instructions are executed by the control processor, the ambient illumination estimation method described in Embodiment 1 is implemented.

[0091] In the computer-readable storage medium, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, digital subscriber line) or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid-state disk (SSD)).

[0092] In summary, the method proposed in this application effectively solves the technical problems of high cost, limited installation location, and single function existing in traditional optical sensors. By presetting the ROI selection algorithm to select multiple ROI regions and combining with the illuminance estimation model for estimation, the environmental illuminance can be estimated more accurately, avoiding the deviation of single-region illuminance estimation, comprehensively considering the illumination characteristics of multiple regions, and improving the illuminance estimation accuracy. By using the ROI selection algorithm to select key regions, the algorithm can adapt to changing environmental illumination conditions and still accurately estimate the illuminance in complex environments. After obtaining the environmental feature data based on the ROI region and then obtaining the final illuminance data through the illuminance estimation model, the obtained illuminance data can be more accurate and reliable. It effectively reduces the hardware cost, avoids the high cost of using traditional optical sensors through the existing camera devices of the vehicle, and realizes multi-functional environmental perception by combining image data and the illuminance estimation model.

[0093] In several embodiments provided in this application, it can be understood that each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code includes one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the block may occur in a different order from that marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved.

[0094] If the described function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing an electronic device to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0095] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present application. It should be understood that the above description is only for the specific embodiments of the present application and is not used to limit the protection scope of the present application. In particular, it is pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. An environmental illuminance estimation method, characterized in that, Including: Real-time collecting image data of the environment where the vehicle is located; Selecting multiple ROI regions from the image data according to a preset ROI selection algorithm; Obtaining environmental feature data based on the ROI regions; And estimating the final environmental illuminance data based on the environmental feature data through a preset illuminance estimation model.

2. The environmental light intensity estimation method according to claim 1, wherein After the real-time collecting of the image data of the environment where the vehicle is located, it further includes: Performing distortion correction processing and format adjustment on the real-time collected image data to obtain the final image data.

3. The environmental illuminance estimation method according to claim 1, characterized in that The selecting of multiple ROI regions from the image data according to a preset ROI selection algorithm includes: Real-time obtaining the current time information and the current vehicle state information; Based on the current time information and the current vehicle state information, using the preset ROI selection algorithm to determine multiple coordinate ranges according to the final image data as the corresponding ROI regions.

4. The environmental illuminance estimation method according to claim 3, characterized in that The obtaining of environmental feature data based on the ROI regions includes: Obtaining the coordinate information and the region size information of the corresponding ROI regions according to the coordinate ranges, and extracting the corresponding pixel data from the image data based on the coordinate information and the region size information; Converting the pixel data into grayscale image data; Generating a brightness distribution histogram based on the grayscale image data; Obtaining the environmental feature data of the ROI regions based on the brightness distribution histogram.

5. The environmental illuminance estimation method according to claim 3, wherein The illuminance estimation model at least includes one of a machine learning model or a mapping function model; the estimating of the environmental illuminance data through a preset illuminance estimation model includes: Based on the current time information, the current vehicle state information, and the environmental feature data, using a pre-trained illuminance estimation model to estimate the first environmental illuminance data; Based on the first environmental illuminance data, real-time adjusting the estimation parameters of the illuminance estimation model.

6. The environmental illuminance estimation method according to claim 5, characterized in that After the estimating of the first environmental illuminance data, it further includes: Identifying the glare regions from the first environmental illuminance data based on a preset illuminance threshold; Performing a weight reduction process or a shielding process on the first environmental illuminance data of the glare regions to obtain the second environmental illuminance data.

7. The environmental illuminance estimation method according to claim 6, wherein After the obtaining of the second environmental illuminance data, it further includes: Real-time obtaining the current weather information; Performing corresponding adaptation adjustment processing on the second environmental illuminance data based on the current weather information to obtain the final environmental illuminance data; Wherein, the adaptation adjustment processing at least includes denoising, defogging, and / or multi-frame fusion processing.

8. The environmental illuminance estimation method according to claim 7, characterized in that, After the obtaining of the final environmental illuminance data, it further includes: Converting the final environmental illuminance data into a preset standard format and outputting the converted final environmental illuminance data to the in-vehicle application program.

9. A system based on the environmental illuminance estimation method according to any one of claims 1-8, characterized in that, The system includes: A data acquisition module for real-time collecting image data of the environment where the vehicle is located; A region selection module for selecting multiple ROI regions from the image data according to a preset ROI selection algorithm; A feature acquisition module for obtaining environmental feature data based on the ROI regions; And an estimation module for estimating the final environmental illuminance data based on the environmental feature data through a preset illuminance estimation model.

10. A computer-readable storage medium storing computer-executable instructions, characterized in that, When the computer-executable instructions are executed by a control processor, the environmental illuminance estimation method according to any one of claims 1-8 is implemented.

Images