A 360-degree image adjustment method, system and controller

Abstract

A 360-degree image adjustment method, system and controller relate to the field of vehicle driving. In this method, the current light conditions are detected by a light sensor or by analyzing the image data of the first image obtained by a camera, and the target exposure parameters are calculated according to the current light conditions, and the target exposure parameters include shutter speed, aperture size and sensitivity; the camera is adjusted according to the target exposure parameters, and a second image is obtained through the adjusted camera; the color average values ​​of different areas in the second image are calculated, and the white balance parameters are calculated according to the color average values, and the white balance parameters include color temperature, hue and saturation; the second image is color corrected according to the white balance parameters to obtain a third image, and multiple third images are spliced ​​to form the final presented image. The implementation of the technical solution provided by this application improves the imaging quality and viewing experience of 360-degree images.

Description

Technical Field

[0001] The present application relates to the technical field of vehicle driving, and in particular to a 360-degree image adjustment method, system and controller. Background Art

[0002] With the continuous development of autonomous driving technology, 360-degree surround-view cameras have become one of the indispensable safety configurations of modern cars. They can help drivers fully understand the vehicle's surroundings, reduce blind spots, and improve driving safety.

[0003] The existing 360-degree surround monitoring system may cause image quality degradation or even display problems such as distorted screen, vertical lines or black screen when encountering light changes, electromagnetic interference or dirty lenses. Current solutions often require manual intervention to restore to normal, lack of automatic adjustment functions, affecting user experience and system reliability.

[0004] Therefore, a method is needed to automatically adjust the 360-degree image. Summary of the invention

[0005] The present application provides a 360-degree image adjustment method, system and controller, which significantly improve the imaging quality and viewing experience of 360-degree images through adaptive light adjustment, color correction and image stitching technology.

[0006] In a first aspect of the present application, a 360-degree image adjustment method is provided, which is applied to a 360-degree image platform. The method comprises:

[0007] Detecting current light conditions by using a light sensor or analyzing image data of a first image acquired by a camera, and calculating target exposure parameters according to the current light conditions, wherein the target exposure parameters include shutter speed, aperture size, and sensitivity;

[0008] Adjusting the camera according to the target exposure parameter, and acquiring a second image through the adjusted camera;

[0009] Calculating color averages of different areas in the second image, and calculating white balance parameters according to the color averages, the white balance parameters including color temperature, hue, and saturation;

[0010] The second image is color corrected according to the white balance parameter to obtain a third image, and a plurality of the third images are spliced ​​to form a final presented image.

[0011] Optionally, detecting the current light condition by using a light sensor or analyzing the image data of the first image acquired by a camera, and calculating the target exposure parameter according to the current light condition includes:

[0012] Analyzing the brightness distribution of the first image to identify highlight areas and dark areas in the first image;

[0013] Calculating a brightness dynamic range of the first image according to a brightness difference between the highlight area and the dark area, and determining a current light condition according to the brightness dynamic range and a preset pipeline threshold, where the current light condition includes low light, normal light, or overexposure;

[0014] The current light condition is input into a preset machine learning model to obtain a target exposure parameter that matches the current light condition.

[0015] Optionally, calculating the color averages of different regions in the second image and calculating the white balance parameter according to the color averages further includes:

[0016] Dividing the second image into a plurality of preset areas, calculating a first color average value of the three primary colors of red, green and blue in a first target preset area, wherein the first target preset area is any one of the plurality of preset areas;

[0017] Determining a target weight according to the position and size of the first target preset area, and performing weighted averaging on the color averages of the plurality of preset areas based on the target weight and the first color average to obtain a second color average of the second image;

[0018] The white balance parameter is determined according to the second color average value in combination with a preset mapping relationship between color and color temperature, hue, and saturation.

[0019] Optionally, calculating the color averages of different regions in the second image and calculating the white balance parameter according to the color averages further includes:

[0020] Determine whether there is a second target preset area in the second image where the difference between the color average value and the second color average value exceeds a threshold;

[0021] When there is a second target preset area in the second image, color difference is performed using the color average value of the adjacent areas of the second target preset area to correct the color average value of the second target preset area;

[0022] The second color average value is recalculated according to the corrected color average value, and the white balance parameter is calculated according to the re-obtained second color average value.

[0023] Optionally, stitching the plurality of third images together to form a final presented image includes:

[0024] When the third image is distorted, a first correction is performed according to the internal parameters of the camera, wherein the internal parameters include the focal length and the optical center position, and a second correction is performed on the third image after the first correction according to the relative position relationship between the cameras to obtain a fourth image;

[0025] Detecting feature points in the fourth image and calculating descriptors of the feature points;

[0026] Feature points in the plurality of fourth images are matched according to the descriptors to determine a relative position relationship between different fourth images, and the plurality of fourth images are spliced ​​according to the relative position relationship to obtain a final presented image.

[0027] Optionally, the method further includes:

[0028] Preprocessing the second image, and performing obstacle detection on the preprocessed second image according to a preset deep learning model, and outputting the category, position and confidence of the obstacle;

[0029] Marking the dynamic obstacles in the image according to the positions and confidence levels of the obstacles;

[0030] When the confidence level is higher than a threshold, a shadow removal algorithm is determined according to the shape, size and environment of the marked obstacle area, and shadow processing is performed on the obstacle area according to the shadow removal algorithm.

[0031] Optionally, the method further includes:

[0032] Analyzing the motion vector of each pixel in the second image to identify a dynamic obstacle, and recording the motion trajectory and speed of the dynamic obstacle;

[0033] The future position of the dynamic obstacle is predicted according to the motion trajectory and speed, and the dynamic obstacle is marked in advance in subsequent images according to the future position so as to perform continuous shadow processing.

[0034] In a second aspect of the present application, a 360-degree image adjustment system is provided, including a detection module, an adjustment module, a calculation module, and a display module, wherein:

[0035] a detection module configured to detect current light conditions through a light sensor or image data of a first image acquired by an analysis camera, and calculate target exposure parameters according to the current light conditions, wherein the target exposure parameters include shutter speed, aperture size, and sensitivity;

[0036] an adjustment module, configured to adjust the camera according to the target exposure parameter, and acquire a second image through the adjusted camera;

[0037] a calculation module configured to calculate color averages of different areas in the second image, and calculate white balance parameters according to the color averages, wherein the white balance parameters include color temperature, hue and saturation;

[0038] The display module is configured to perform color correction on the second image according to the white balance parameter to obtain a third image, and splice a plurality of the third images to form a final presented image.

[0039] In a third aspect of the present application, a controller is provided, comprising a processor, a memory, a user interface and a network interface, wherein the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the controller executes any of the methods described above.

[0040] In a fourth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores instructions, and when the instructions are executed, any of the methods described above is executed.

[0041] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0042] 1. Obtain the current light conditions through the light sensor or analysis camera, and calculate the target exposure parameters (shutter speed, aperture size and sensitivity) accordingly, which can automatically adapt to different light environments. Whether it is strong light or weak light conditions, it can ensure that the camera captures clear and moderately bright images, thereby improving the imaging quality of 360 images;

[0043] 2. By calculating the color averages of different areas in the second image and calculating the white balance parameters (color temperature, hue and saturation) based on these color averages, accurate correction of the image color can be achieved. Adjustment of white balance is crucial to ensure the authenticity and accuracy of image color, especially under complex lighting conditions. This step can significantly improve the color performance of 360 images;

[0044] 3. Multiple color-corrected third images are stitched together to form the final 360-degree panoramic image. Since each image has been optimized in color and brightness, the stitched image is more visually coherent and consistent, reducing stitching marks caused by light and color differences and improving the overall viewing experience;

[0045] 4. The entire adjustment process is highly automated, relying on the calculation results of sensors and algorithms, reducing the need for manual intervention. This not only improves work efficiency, but also makes the adjustment of 360 images more intelligent and precise. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flowchart of a 360-degree image adjustment method disclosed in an embodiment of the present application;

[0047] Figure 2 It is a module schematic diagram of a 360-degree image adjustment system disclosed in an embodiment of the present application;

[0048] Figure 3 It is a structural schematic diagram of a controller disclosed in an embodiment of the present application.

[0049] Explanation of the reference numerals: 201, detection module; 202, adjustment module; 203, calculation module; 204, display module; 301, processor; 302, communication bus; 303, user interface; 304, network interface; 305, memory. DETAILED DESCRIPTION

[0050] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments.

[0051] In the description of the embodiments of the present application, words such as "for example" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "for example" or "for example" is intended to present related concepts in a specific way.

[0052] In the description of the embodiments of the present application, the meaning of the term "multiple" refers to two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. The terms "include", "comprise", "have" and their variations all mean "including but not limited to", unless otherwise specifically emphasized.

[0053] This embodiment discloses a 360-degree image adjustment method, which is applied to a 360-degree image platform. Figure 1 is a flow chart of the 360-degree image adjustment method disclosed in the embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:

[0054] S110, detecting current light conditions by using a light sensor or analyzing image data of a first image acquired by a camera, and calculating target exposure parameters according to the current light conditions, wherein the target exposure parameters include shutter speed, aperture size, and sensitivity;

[0055] The onboard 360-degree imaging platform captures real-time images of the vehicle's surroundings through multiple high-definition cameras installed on the front, rear and side mirrors of the vehicle. These images are processed by the on-board computer and seamlessly stitched into a 360-degree panoramic bird's-eye view, which is displayed on the infotainment display on the vehicle's center console. The driver can intuitively view the environment around the vehicle through this display, thereby eliminating blind spots and improving driving safety. A light sensor is a device that can sense the intensity of surrounding light. In the 360-degree imaging platform, light sensors are used to detect current light conditions in real time. The image data captured by the camera contains rich light information, and by analyzing these image data, the current light conditions can also be inferred. Light conditions include light intensity and light direction. Exposure parameters refer to settings that affect the camera's exposure, including shutter speed, aperture size, and sensitivity. These parameters together determine the camera's ability to capture light, thereby affecting the brightness and contrast of the image. The shutter speed determines the length of time the camera lens is open, which affects the amount of light entering the camera. In the case of sufficient light, the shutter speed can be set faster to reduce the amount of light entering and avoid overexposure of the image. In low light conditions, the shutter speed needs to be set slower to increase the amount of light entering and ensure that the image is bright enough. The aperture size determines the size of the opening in the camera lens that allows light to pass through. The larger the aperture, the more light enters the camera; the smaller the aperture, the less light enters the camera. Based on the current light conditions, the system automatically adjusts the aperture size to achieve the best exposure effect. Sensitivity refers to the sensitivity of the camera sensor to light. In low light conditions, the sensitivity needs to be increased to increase the sensor's ability to capture light. In sufficient light, the sensitivity can be reduced to avoid noise in the image. Based on the calculated target exposure parameters, the system automatically adjusts the camera. These adjustments include changing settings such as shutter speed, aperture size, and sensitivity to ensure that the camera can capture the best image. By accurately calculating and adjusting exposure parameters, the clarity, contrast, and color performance of the image can be significantly improved. This is particularly important for 360 imaging platforms, because high-quality images can bring users a more realistic and immersive visual experience.

[0056] Optionally, detecting the current light condition by using a light sensor or analyzing the image data of the first image acquired by a camera, and calculating the target exposure parameter according to the current light condition includes:

[0057] Analyzing the brightness distribution of the first image to identify highlight areas and dark areas in the first image;

[0058] Calculating a brightness dynamic range of the first image according to a brightness difference between the highlight area and the dark area, and determining a current light condition according to the brightness dynamic range and a preset pipeline threshold, where the current light condition includes low light, normal light, or overexposure;

[0059] The current light condition is input into a preset machine learning model to obtain a target exposure parameter that matches the current light condition.

[0060] The system will conduct a detailed analysis of the brightness distribution of the first image captured by the camera. Brightness distribution refers to the distribution of brightness values ​​of each pixel in the image, which reflects the intensity and distribution of light in the image. Based on the brightness distribution analysis, the system will further identify the highlight area and dark area in the image. The highlight area usually refers to the part of the image with higher brightness value, while the dark area refers to the part with lower brightness value. The identification of these areas helps the system understand the current light conditions more accurately. The system will calculate the brightness dynamic range of the image. The brightness dynamic range refers to the brightness difference between the brightest and darkest areas in the image, which reflects the range and contrast of light in the image. By comparing the brightness dynamic range with the preset light threshold, the system can determine whether the current light condition is low light, normal light or overexposure. Once the current light condition is determined, the system will input this information into a preset machine learning model. This machine learning model has been trained to output target exposure parameters that match the input light conditions. The target exposure parameters include shutter speed, aperture size, and sensitivity, which together determine the camera's ability to capture light and the exposure effect of the image. The system will adjust the camera accordingly based on the target exposure parameters output by the machine learning model. These adjustments are designed to ensure that the camera can capture the best image effects under the current lighting conditions. At the same time, the system will also provide real-time feedback and adjustments based on the actual shooting results to ensure that the image clarity and color performance are always in the best state.

[0061] By accurately analyzing the brightness distribution of the image and identifying the highlight and dark areas, the current light conditions can be judged more accurately. On this basis, the optimal exposure parameters that match the target light conditions are calculated to ensure that the image captured by the camera can reach the best state in terms of brightness, contrast and color. In the case of insufficient light or excessive light, the image is prone to noise. By automatically adjusting the exposure parameters, the generation of noise can be reduced, making the image clearer and more delicate. Applying this technology in the in-vehicle 360 ​​imaging platform can ensure that the camera can capture clear images under different light conditions. This helps the driver to understand the environment around the vehicle more accurately and improve driving safety. In the case of insufficient light or excessive light, the driver's vision is easily disturbed, thereby increasing the risk of accidents. By automatically adjusting the exposure parameters, this technology can reduce this interference and reduce the possibility of accidents. The exposure parameters can be automatically adjusted according to the current light conditions without manual intervention. This reflects the intelligence level of the in-vehicle system and improves the convenience and ease of use of the system.

[0062] S120, adjusting the camera according to the target exposure parameter, and acquiring a second image through the adjusted camera;

[0063] The system sends the target exposure parameters to the camera, which automatically adjusts its internal mechanical structure and electronic components to match these parameters. For example, if the target exposure parameters require a faster shutter speed, the camera will shorten the shutter opening time accordingly; if a larger aperture is required, the camera will adjust the aperture opening degree. The adjusted camera begins to capture a new image, the second image. Because the camera has been optimized according to the target exposure parameters, the captured second image will be more accurate and vivid in brightness, contrast and color. Compared with the unadjusted image (i.e., the first image), the second image will be visually improved. The highlight areas will not be overexposed, and more details will be preserved in the dark areas. The system also analyzes the captured second image in real time to evaluate the effect of the adjustment. If the image quality is still not ideal, the system will fine-tune the exposure parameters based on the analysis results. Through this real-time feedback and adjustment mechanism, the system can continuously optimize the camera's exposure settings. This helps ensure that the camera can capture the best image effects in different lighting conditions. The captured image is clearer, which helps the driver to more accurately understand the environment around the vehicle. This improves driving safety and reduces the risk of accidents caused by unclear vision. The colors of the images are more realistic and natural, improving the overall visual effect of the in-vehicle 360 ​​imaging platform.

[0064] S130, calculating color averages of different regions in the second image, and calculating white balance parameters according to the color averages, where the white balance parameters include color temperature, hue, and saturation;

[0065] The second image is divided into a plurality of regions, which may be fixed grid-like divisions or dynamic divisions based on the image content (e.g., based on the boundaries of objects in the image). For each region, the system extracts its color information, which typically includes the values ​​of the three channels of red, green, and blue (RGB). The system calculates the color average of all pixels in each region to obtain the average RGB value of each region. Color temperature is a measure of the color of a light source, usually in Kelvin (K). The system estimates the color temperature of the current light by analyzing the color averages of different regions, especially the ratio of the red to blue channels. If the value of the red channel is relatively high and the value of the blue channel is relatively low, it may indicate a low color temperature (warmer tones); conversely, it may indicate a high color temperature (cooler tones). Hue refers to the overall tendency of the color in the image, such as reddish, greenish, or bluish. The system calculates the hue offset of the image by comparing the color averages of different regions with the standard color values. Based on the offset, the system can adjust the hue of the image to make it closer to natural colors. Saturation refers to the vividness of the colors in the image. The system estimates the saturation of the image by analyzing the RGB ratios in the color averages and the differences between these ratios and the standard color values. If the color average of a region deviates far from the standard color value and the RGB ratios differ greatly, it may indicate that the saturation of the region is high; conversely, it may indicate low saturation. The system applies the calculated color temperature, hue, and saturation parameters to the second image to adjust its white balance. This usually involves transforming the RGB values ​​of each pixel in the image to make it consistent with the target color temperature, hue, and saturation. The system evaluates the effect of the adjusted image to ensure that the color of the image after the white balance adjustment is more accurate and natural. If the evaluation result is not ideal, the system may make fine adjustments based on the feedback. Through white balance adjustment, the color of the image is more accurate and natural, which conforms to the visual habits of the human eye. This helps to improve the driver's visual experience and reduce misunderstandings or confusion caused by color distortion. The adjusted image color is softer and more natural, which helps to reduce the driver's visual fatigue. This is especially important for drivers who drive for a long time, which can improve their driving safety and comfort.

[0066] Optionally, calculating the color averages of different regions in the second image and calculating the white balance parameter according to the color averages further includes:

[0067] Dividing the second image into a plurality of preset areas, calculating a first color average value of the three primary colors of red, green and blue in a first target preset area, wherein the first target preset area is any one of the plurality of preset areas;

[0068] Determining a target weight according to the position and size of the first target preset area, and performing weighted averaging on the color averages of the plurality of preset areas based on the target weight and the first color average to obtain a second color average of the second image;

[0069] The white balance parameter is determined according to the second color average value in combination with a preset mapping relationship between color and color temperature, hue, and saturation.

[0070] The system first divides the second image into a plurality of preset areas. These preset areas can be fixed, such as grid-shaped divisions, or dynamic, divided according to the image content or specific needs. For each preset area, the system calculates the color averages of its three primary colors of red, green, and blue. These averages reflect the overall distribution of colors in the area. The system determines its target weight according to the position and size of the first target preset area. The size of the weight may depend on the importance of the area in the image, such as the center area may have a higher weight than the edge area. Based on the target weight and the color average of the first target preset area, the system performs a weighted average of the color averages of the plurality of preset areas. The purpose of this step is to comprehensively consider the color information of different areas to obtain the color average (i.e., the second color average) of the entire image. The system has preset mapping relationships between colors and color temperature, hue, and saturation. These mapping relationships are based on the principles of color science and associate color averages with specific white balance parameters (color temperature, hue, and saturation). Based on the calculated second color averages, the system determines the white balance parameters in combination with the preset mapping relationships. These parameters will be used to adjust the color of the image to make it more accurate and natural. Through the above steps, the system can calculate the white balance parameters more accurately, thereby adjusting the color of the image to make it closer to the color of the real world. The color of the image after white balance adjustment is more natural and vivid, which helps to improve the overall quality of the image.

[0071] By dividing the second image into a plurality of preset areas and calculating the color average of each area, especially performing a fine calculation for the first target preset area, the color distribution of different areas in the image can be more accurately reflected. Subsequently, the target weight is determined according to the position and size of the area, and the weighted average is performed to obtain the second color average of the entire image. This process effectively avoids the influence of the color of a single area on the overall color and improves the accuracy of color calculation. Based on the second color average, combined with the preset mapping relationship between color and color temperature, hue, and saturation, the system can determine more accurate white balance parameters. These parameters can more accurately adjust the color of the image to make it closer to the color of the real world, thereby avoiding the occurrence of color distortion or color cast. After the white balance parameters are adjusted, the image has more natural and vivid colors, and the contrast and brightness are more moderate. This not only improves the overall quality of the image, but also makes the details in the image clearer and more discernible, providing the driver with clearer and more accurate visual information. Applying the above method in the vehicle-mounted 360 imaging platform can enable the driver to obtain a more realistic and natural color experience when viewing the surrounding environment. This helps the driver to judge the distance and obstacles more accurately and improve driving safety. At the same time, images with accurate colors can also help reduce the driver's visual fatigue and improve driving comfort. Not only is it suitable for different lighting conditions and camera characteristics, it can also be fine-tuned according to actual needs. For example, the division of preset areas, the determination of target weights, and the selection of mapping relationships can be adjusted according to different car models, different driving environments, or different driver preferences. This makes the method more flexible and adaptable, and can better meet the needs of different users.

[0072] Optionally, calculating the color averages of different regions in the second image and calculating the white balance parameter according to the color averages further includes:

[0073] Determine whether there is a second target preset area in the second image where the difference between the color average value and the second color average value exceeds a threshold;

[0074] When there is a second target preset area in the second image, color difference is performed using the color average value of the adjacent areas of the second target preset area to correct the color average value of the second target preset area;

[0075] The second color average value is recalculated according to the corrected color average value, and the white balance parameter is calculated according to the re-obtained second color average value.

[0076] First, a color difference threshold is set. This threshold is used to determine whether the color average value of a certain area is significantly different from the overall color average value (i.e., the second color average value). The system compares the difference between the color average value of each preset area in the second image and the second color average value. If the difference between the color average value of a certain area and the second color average value exceeds the set threshold, the area is determined to be a color abnormal area, i.e., the second target preset area. For the detected color abnormal area (the second target preset area), the system selects the color average value of its adjacent areas for color difference correction. The selection of adjacent areas can be made according to actual conditions, such as selecting directly adjacent areas, areas within a certain range around, etc. The system uses the color average value of adjacent areas to correct the color average value of the color abnormal area through a certain algorithm (such as weighted averaging, interpolation, etc.). The purpose of correction is to make the color of the color abnormal area closer to the overall color, thereby reducing the color difference. After correcting the color abnormal area, the system needs to recalculate the second color average value of the entire image. This step takes into account the color average value of the corrected color abnormal area, making the second color average value more accurate. The system recalculates the white balance parameters based on the re-obtained second color average value, combined with the preset mapping relationship between color and color temperature, hue, and saturation. These parameters will be used to adjust the color of the image to make it more accurate and natural. By detecting and correcting color abnormal areas, the system can more accurately reflect the color distribution of different areas in the image and improve the accuracy of color calculation. The corrected image color is more natural and uniform, avoiding color distortion or color cast, thereby improving the overall quality of the image. Color-accurate images help drivers judge distances and obstacles more accurately and improve driving safety. At the same time, it also reduces the driver's visual fatigue and improves driving comfort. The ability to process color abnormal areas allows the system to maintain good color accuracy under different lighting conditions, camera characteristics or image content, enhancing the robustness of the system.

[0077] When there is an area (i.e., the second target preset area) in the second image where the difference between the color average value and the overall color average value (i.e., the second color average value) exceeds a set threshold, the color difference correction is performed using the color average value of the adjacent areas of the area. This method can more accurately identify and correct the color deviation area in the image, so that the corrected color average value is closer to the real color, thereby improving the accuracy of color correction. By performing color correction on the second target preset area, the color of the area can be made more coherent and consistent with the color of the adjacent areas. This helps to reduce color mutations and distortion in the image and improve the overall visual effect of the image. In particular, in the in-vehicle 360 ​​imaging platform, coherent and consistent color performance can enhance the driver's perception and understanding of the surrounding environment and improve driving safety. Recalculating the second color average value based on the corrected color average value and calculating the white balance parameters based on this can make the white balance parameters more adaptable to the actual color distribution of the image. This improved adaptability helps to ensure that the image can present natural and real color effects under different lighting conditions and camera characteristics. Images with accurate, coherent and consistent colors can provide users with a better visual experience. In the in-vehicle 360-degree imaging platform, the optimized color performance can make the driver more comfortable and pleasant when viewing the surrounding environment, reducing visual fatigue and misunderstanding. At the same time, this can also improve the overall satisfaction and trust of users in the in-vehicle imaging system.

[0078] Optionally, the method further includes:

[0079] Preprocessing the second image, and performing obstacle detection on the preprocessed second image according to a preset deep learning model, and outputting the category, position and confidence of the obstacle;

[0080] Marking the dynamic obstacles in the image according to the positions and confidence levels of the obstacles;

[0081] When the confidence level is higher than a threshold, a shadow removal algorithm is determined according to the shape, size and environment of the marked obstacle area, and shadow processing is performed on the obstacle area according to the shadow removal algorithm.

[0082] In the in-vehicle 360 ​​imaging platform, the original image may be affected by many factors, such as lighting conditions, camera characteristics, noise, etc. Therefore, before obstacle detection, the second image needs to be preprocessed to improve the quality and clarity of the image. The preprocessing steps may include denoising, contrast enhancement, brightness adjustment, etc. to ensure that the details and features in the image can be clearly presented. The preprocessed image is input into a preset deep learning model. This deep learning model has been trained to identify and detect obstacles in the image. The model outputs the category of the obstacle (such as pedestrians, vehicles, animals, etc.), location (such as coordinates, bounding box, etc.) and confidence (that is, the model's confidence in the detection result). This information is crucial for subsequent dynamic obstacle marking and shadow processing. Based on the location and confidence of the obstacle, the system can mark the dynamic obstacles in the image. Dynamic obstacles are those that may move or change, such as pedestrians, vehicles, etc. The marking process may involve drawing bounding boxes, highlighting, or adding labels on the image so that the driver can clearly see these obstacles. When the confidence level of an obstacle is higher than the set threshold, the system considers the obstacle detection result to be reliable. At this point, the system will determine the appropriate shadow removal algorithm based on the shape, size and environment of the marked obstacle area. The purpose of the shadow removal algorithm is to reduce or eliminate the shadows in the obstacle area to improve the clarity and accuracy of the image. The system needs to identify the shadow part in the obstacle area. This can be achieved by analyzing the brightness, color and other features in the image. When the shadow part is identified, the system can apply the corresponding shadow removal algorithm to reduce or eliminate the shadow. These algorithms may involve image restoration, interpolation, filtering and other technologies. The system will optimize the processed image to ensure that the area after shadow removal is consistent with the color and brightness of the surrounding environment.

[0083] By preprocessing the second image, such as denoising and contrast enhancement, the image quality can be significantly improved, so that the deep learning model can more accurately identify and detect obstacles in the image. At the same time, the application of the deep learning model also further improves the accuracy of obstacle detection, and can output the category, location and confidence of the obstacle, providing a reliable basis for subsequent dynamic obstacle marking and shadow processing. Marking dynamic obstacles in the image according to the location and confidence of the obstacle can help the driver see the surrounding obstacles more clearly, especially those that may move or change, such as pedestrians and vehicles. This helps the driver better perceive and understand the surrounding environment and improve driving safety. When the confidence of the obstacle is higher than the set threshold, the system will determine the appropriate shadow removal algorithm based on the shape, size and environment of the marked obstacle area, and perform shadow processing on the obstacle area. This step can significantly reduce or eliminate the shadows in the obstacle area and improve the clarity and accuracy of the image. The optimized image quality can provide a better visual experience, and also help the driver to more accurately judge the distance and location of the obstacle. The application of the above method not only improves the practicality of the vehicle-mounted 360 imaging platform, but also improves the intelligence level of the system. Through the application of deep learning models and the combination of image processing technology, the system can automatically detect, mark and handle obstacles, reducing the need for manual intervention and improving the automation and intelligence of the system. The optimized in-vehicle 360 ​​imaging platform can provide clearer and more accurate image information, helping drivers to better perceive and understand the surrounding environment. This not only improves driving safety, but also enhances the user experience. Drivers can more easily view surrounding obstacles and road conditions, reducing tension and fatigue during driving.

[0084] Optionally, the method further includes:

[0085] Analyzing the motion vector of each pixel in the second image to identify a dynamic obstacle, and recording the motion trajectory and speed of the dynamic obstacle;

[0086] The future position of the dynamic obstacle is predicted according to the motion trajectory and speed, and the dynamic obstacle is marked in advance in subsequent images according to the future position so as to perform continuous shadow processing.

[0087] The on-board 360-degree imaging platform first captures consecutive image frames (i.e., the second image and its subsequent images). For each frame of the image, the system analyzes the motion vector of each pixel in it. The motion vector refers to the displacement of a pixel between consecutive frames, which can be calculated by algorithms such as optical flow and block matching. By comparing the position changes of pixels in adjacent frames, the system can identify pixels belonging to dynamic obstacles (such as vehicles, pedestrians, etc.) and distinguish the difference between them and the static background. When dynamic obstacles are identified, the system tracks the movement of these obstacles and records their movement trajectories (i.e., a series of coordinate points that change over time). At the same time, the system also calculates the speed of dynamic obstacles, which can be achieved by measuring their displacement and time difference between consecutive frames. Based on the movement trajectory and speed of dynamic obstacles, the system can use prediction algorithms (such as linear prediction, Kalman filtering, etc.) to estimate their positions in future frames. This prediction capability is crucial for marking dynamic obstacles in advance because it allows the system to process them before they actually appear. In subsequent image frames, the system marks dynamic obstacles in advance based on the predicted future positions. This can be achieved by drawing bounding boxes, highlighting, or adding labels on the image. Early marking not only helps the driver to detect and identify dynamic obstacles earlier, but also facilitates subsequent shadow processing. Once the dynamic obstacles are marked, the system can apply the shadow removal algorithm mentioned earlier to shadow the obstacle area. Since the system can predict and mark dynamic obstacles in advance, continuous shadow processing of these obstacles can be achieved, even if they move or change shape in the image. This continuous shadow processing ensures that dynamic obstacles in the image are always clear and accurate, thereby improving driving safety and user experience.

[0088] By analyzing the motion vector of each pixel, this method can accurately capture the motion characteristics of dynamic obstacles, effectively distinguish static and dynamic objects, and improve the accuracy of recognition. The motion trajectory and speed of dynamic obstacles are recorded in real time to achieve real-time tracking of dynamic obstacles, providing reliable data support for subsequent prediction and marking. According to the motion trajectory and speed of dynamic obstacles, their future positions can be predicted, so that they can be marked in advance in subsequent images to ensure continuous processing of dynamic obstacles. Continuous shadow processing can maintain the stability of dynamic obstacle processing and avoid processing interruptions or error accumulation caused by the movement of dynamic obstacles. By marking dynamic obstacles in advance, the system can provide drivers with earlier warning information to help drivers react in advance and avoid potential safety hazards. Continuous shadow processing can improve image quality, make dynamic obstacles clearer and more prominent in the image, and enhance the driver's visual perception ability. The application of this method makes the vehicle-mounted imaging system more intelligent and humanized, providing drivers with clearer and more intuitive driving environment information and improving driving experience. This method combines multiple technical means such as motion vector analysis, trajectory prediction, and shadow processing, providing new ideas and methods for the technical innovation of vehicle-mounted imaging systems. With the continuous advancement of technology and the continuous expansion of application scenarios, this method is expected to promote the development of vehicle imaging systems towards a more intelligent and automated direction.

[0089] S140: Perform color correction on the second image according to the white balance parameter to obtain a third image, and splice a plurality of the third images to form a final presented image.

[0090] The color of the second image is adjusted using the previously calculated white balance parameters. White balance parameters usually include gain values ​​for the red, green, and blue channels, which are used to correct color deviations in the image to make it closer to the colors observed by the human eye under natural light. When applying white balance parameters, each pixel in the image will be adjusted for gain according to the value of its color channel. For example, if the image is reddish overall, the gain value of the red channel may be reduced while the gain values ​​of the green and blue channels may be increased to achieve a color balance effect. The corrected image (i.e., the third image) will be more accurate and natural in color, avoiding color distortion caused by changes in light sources or camera characteristics. In the in-vehicle 360 ​​imaging platform, it is usually necessary to stitch images from different cameras (such as front, rear, left, right, etc.) to form a complete surround image. These images have been color-corrected before stitching to ensure color consistency. Image stitching involves multiple technical challenges, such as image registration (i.e., determining the relative position relationship between different images) and image fusion (i.e., smoothly transitioning the boundaries between different images). In this step, image processing algorithms are used to automatically identify and match feature points between different images to determine their relative positional relationships. Image fusion technology is then used to smoothly transition the boundaries of these images to avoid obvious stitching marks. After the stitching process, a complete and seamless surround view image is obtained. This image is not only accurate and natural in color, but also provides an all-round perspective to help drivers better perceive and understand the surrounding driving environment.

[0091] Optionally, stitching the plurality of third images together to form a final presented image includes:

[0092] When the third image is distorted, a first correction is performed according to the internal parameters of the camera, wherein the internal parameters include the focal length and the optical center position, and a second correction is performed on the third image after the first correction according to the relative position relationship between the cameras to obtain a fourth image;

[0093] Detecting feature points in the fourth image and calculating descriptors of the feature points;

[0094] Feature points in the plurality of fourth images are matched according to the descriptors to determine a relative position relationship between different fourth images, and the plurality of fourth images are spliced ​​according to the relative position relationship to obtain a final presented image.

[0095] In the vehicle-mounted 360 imaging platform, due to the lens characteristics of the camera, the captured image may be distorted, such as barrel distortion or pincushion distortion. In order to eliminate this distortion, the third image needs to be first corrected according to the intrinsic parameters of the camera (including focal length and optical center position). Intrinsic parameters are parameters inherent to the camera, which describe the imaging characteristics of the camera. By applying these parameters, the image can be distorted to make it closer to the shape of the real world. In addition to distortion correction, the third image after the first correction needs to be second corrected according to the relative position relationship between the cameras. This is because in actual applications, different cameras may be located at different positions of the vehicle, and the images they capture will differ in perspective and position. By measuring or calibrating the relative position relationship between the cameras (such as rotation matrix and translation vector), the first corrected images can be further transformed to ensure that they can be correctly aligned when stitched. In the image stitching process, feature points are key points that connect different images. They are usually significant structures or edges in the image, such as corner points, edge intersections, etc. Use feature point detection algorithms (such as SIFT, SURF, ORB, etc.) to detect feature points in the fourth image and calculate their descriptors. A descriptor is a vector that represents the image information around the feature point, which is used to match the feature points in subsequent steps. With the descriptors of the feature points, a descriptor matching algorithm can be used to find similar feature points between different fourth images. These matched feature points will be used as reference points when stitching images. Matching algorithms are usually based on similarity metrics between descriptors (such as Euclidean distance, Hamming distance, etc.), and apply some strategies (such as nearest neighbor search, ratio test, etc.) to screen reliable matching pairs. Through the matched feature points, the relative position relationship between different fourth images can be calculated. This usually involves solving a transformation matrix (such as a homography matrix or an affine transformation matrix) that describes the transformation relationship from one image to another. With the relative position relationship, multiple fourth images can be stitched together using an image stitching algorithm. Stitching algorithms usually involve steps such as image transformation and image fusion to ensure that the stitched image is visually seamless and natural. In the stitching process, some technical challenges need to be dealt with, such as brightness differences between images, color inconsistency, stitching gaps, etc. These problems can be effectively solved by applying appropriate image processing techniques (such as histogram matching, multi-band fusion, etc.).

[0096] By performing the first correction based on the camera's intrinsic parameters (focal length and optical center position), image distortion caused by camera lens distortion can be effectively eliminated. This step is the basis for ensuring the accuracy and quality of subsequent image stitching. Further, the second correction of the image after the first correction based on the relative position relationship between the cameras can further adjust the relative position and angle between the images so that they are more accurately aligned. By detecting the feature points in the image and calculating their descriptors, representative local features in the image can be extracted. These features play a key role in the image stitching process because they can establish a reliable matching relationship between different images. By matching the feature points in multiple images according to the descriptors, the relative position relationship between different images can be determined. The accuracy and stability of this step directly affects the quality of the final stitched image. The entire stitching process is highly automated, from distortion correction to feature point detection and matching, to the final image stitching, all relying on advanced image processing algorithms and computing techniques. This greatly improves the efficiency and accuracy of image stitching. After precise distortion correction and feature point matching, multiple images can be seamlessly stitched together to form a complete and coherent final presentation image. The stitched images maintain consistency in color, brightness, contrast, etc., presenting high-quality, high-definition visual effects.

[0097] This embodiment also discloses a 360-degree image adjustment system. Figure 2 is a schematic diagram of a module of a 360-degree image adjustment system disclosed in an embodiment of the present application, such as Figure 2 As shown, the system includes a detection module 201, an adjustment module 202, a calculation module 203 and a display module 204, wherein:

[0098] The detection module 201 is configured to detect the current light condition through the light sensor or the image data of the first image acquired by the analysis camera, and calculate the target exposure parameters according to the current light condition, wherein the target exposure parameters include shutter speed, aperture size and sensitivity;

[0099] An adjustment module 202, configured to adjust the camera according to the target exposure parameter, and acquire a second image through the adjusted camera;

[0100] A calculation module 203, configured to calculate the color average values ​​of different regions in the second image, and calculate white balance parameters according to the color average values, wherein the white balance parameters include color temperature, hue and saturation;

[0101] The display module 204 is configured to perform color correction on the second image according to the white balance parameter to obtain a third image, and splice a plurality of the third images to form a final displayed image.

[0102] Optionally, the detection module 201 is configured to:

[0103] Analyzing the brightness distribution of the first image to identify highlight areas and dark areas in the first image;

[0104] Calculating a brightness dynamic range of the first image according to a brightness difference between the highlight area and the dark area, and determining a current light condition according to the brightness dynamic range and a preset pipeline threshold, where the current light condition includes low light, normal light, or overexposure;

[0105] The current light condition is input into a preset machine learning model to obtain a target exposure parameter that matches the current light condition.

[0106] Optionally, the calculation module 203 is configured to:

[0107] Dividing the second image into a plurality of preset areas, calculating a first color average value of the three primary colors of red, green and blue in a first target preset area, wherein the first target preset area is any one of the plurality of preset areas;

[0108] Determining a target weight according to the position and size of the first target preset area, and performing weighted averaging on the color averages of the plurality of preset areas based on the target weight and the first color average to obtain a second color average of the second image;

[0109] The white balance parameter is determined according to the second color average value in combination with a preset mapping relationship between color and color temperature, hue, and saturation.

[0110] Optionally, the calculation module 203 is further configured to:

[0111] Determine whether there is a second target preset area in the second image where the difference between the color average value and the second color average value exceeds a threshold;

[0112] When there is a second target preset area in the second image, color difference is performed using the color average value of the adjacent areas of the second target preset area to correct the color average value of the second target preset area;

[0113] The second color average value is recalculated according to the corrected color average value, and the white balance parameter is calculated according to the re-obtained second color average value.

[0114] Optionally, the display module 204 is configured to:

[0115] When the third image is distorted, a first correction is performed according to the internal parameters of the camera, wherein the internal parameters include the focal length and the optical center position, and a second correction is performed on the third image after the first correction according to the relative position relationship between the cameras to obtain a fourth image;

[0116] Detecting feature points in the fourth image and calculating descriptors of the feature points;

[0117] Feature points in the plurality of fourth images are matched according to the descriptors to determine a relative position relationship between different fourth images, and the plurality of fourth images are spliced ​​according to the relative position relationship to obtain a final presented image.

[0118] Optionally, the system further comprises a shadow module, wherein the shadow module is configured to:

[0119] Preprocessing the second image, and performing obstacle detection on the preprocessed second image according to a preset deep learning model, and outputting the category, position and confidence of the obstacle;

[0120] Marking the dynamic obstacles in the image according to the positions and confidence levels of the obstacles;

[0121] When the confidence level is higher than a threshold, a shadow removal algorithm is determined according to the shape, size and environment of the marked obstacle area, and shadow processing is performed on the obstacle area according to the shadow removal algorithm.

[0122] Optionally, the system further comprises a marking module, wherein the marking module is configured to:

[0123] Analyzing the motion vector of each pixel in the second image to identify a dynamic obstacle, and recording the motion trajectory and speed of the dynamic obstacle;

[0124] The future position of the dynamic obstacle is predicted according to the motion trajectory and speed, and the dynamic obstacle is marked in advance in subsequent images according to the future position so as to perform continuous shadow processing.

[0125] It should be noted that: when the device provided in the above embodiment realizes its function, only the division of the above functional modules is used as an example. In actual application, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0126] This embodiment also discloses a controller, referring to Figure 3The controller may include: at least one processor 301 , at least one communication bus 302 , a user interface 303 , a network interface 304 , and at least one memory 305 .

[0127] The communication bus 302 is used to realize the connection and communication between these components.

[0128] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0129] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0130] Among them, the processor 301 may include one or more processing cores. The processor 301 uses various interfaces and lines to connect various parts in the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 305, and calling data stored in the memory 305. Optionally, the processor 301 can be implemented in at least one hardware form of digital signal processing (Digital Signal Processing, DSP), field programmable gate array (Field-Programmable Gate Array, FPGA), and programmable logic array (Programmable Logic Array, PLA). The processor 301 can integrate one or a combination of a central processing unit (Central Processing Unit, CPU), a graphics processing unit (Graphics Processing Unit, GPU) and a modem. Among them, the CPU mainly processes the operating system, user interface and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 301, and it can be implemented separately through a chip.

[0131] Among them, the memory 305 may include a random access memory (Random Access Memory, RAM) and may also include a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be optionally at least one storage device located away from the aforementioned processor 301. As Figure 3 As shown, the memory 305 as a computer storage medium may include an operating system, a network communication module, a user interface 303 module, and an application program for a 360-degree image adjustment method.

[0132] exist Figure 3 In the controller shown, the user interface 303 is mainly used to provide an input interface for the user and obtain data input by the user; and the processor 301 can be used to call the application program for the adjustment method of the 360 ​​image stored in the memory 305. When executed by one or more processors 301, the controller executes one or more methods in the above-mentioned embodiments.

[0133] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required for the present application.

[0134] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0135] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some service interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0136] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0137] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0138] If 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 memory. Based on this understanding, the technical solution of the present application, 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, which is stored in a memory 305 and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned memory 305 includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk.

[0139] The above is only an exemplary embodiment of the present disclosure and cannot be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure are still within the scope of the present disclosure. After considering the disclosure of the specification, those skilled in the art will easily think of other embodiments of the present disclosure. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary technical means in the technical field that are not recorded in the present disclosure. The description and examples are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A 360-degree image adjustment method, characterized in that: Applied to a 360-degree imaging platform, the method includes: Detecting current light conditions by using a light sensor or analyzing image data of a first image acquired by a camera, and calculating target exposure parameters according to the current light conditions, wherein the target exposure parameters include shutter speed, aperture size, and sensitivity; Adjusting the camera according to the target exposure parameter, and acquiring a second image through the adjusted camera; Calculating color averages of different areas in the second image, and calculating white balance parameters according to the color averages, the white balance parameters including color temperature, hue, and saturation; Performing color correction on the second image according to the white balance parameter to obtain a third image, and splicing a plurality of the third images to form a final presented image, The step of calculating the color average values ​​of different regions in the second image and calculating the white balance parameter according to the color average values ​​further includes: Dividing the second image into a plurality of preset areas, calculating a first color average value of the three primary colors of red, green and blue in a first target preset area, wherein the first target preset area is any one of the plurality of preset areas; Determining a target weight according to the position and size of the first target preset area, and performing weighted averaging on the color averages of the plurality of preset areas based on the target weight and the first color average to obtain a second color average of the second image; The white balance parameter is determined according to the second color average value in combination with a preset mapping relationship between color and color temperature, hue, and saturation.

2. The 360-degree image adjustment method according to claim 1, characterized in that: The detecting current light conditions by using a light sensor or analyzing the image data of the first image acquired by a camera, and calculating the target exposure parameter according to the current light conditions comprises: Analyzing the brightness distribution of the first image to identify highlight areas and dark areas in the first image; Calculating a brightness dynamic range of the first image according to a brightness difference between the highlight area and the dark area, and determining a current light condition according to the brightness dynamic range and a preset pipeline threshold, where the current light condition includes low light, normal light, or overexposure; The current light condition is input into a preset machine learning model to obtain a target exposure parameter that matches the current light condition.

3. The 360-degree image adjustment method according to claim 1, characterized in that: The step of calculating the color average values ​​of different regions in the second image and calculating the white balance parameter according to the color average values ​​further includes: Determine whether there is a second target preset area in the second image where the difference between the color average value and the second color average value exceeds a threshold; When there is a second target preset area in the second image, color difference is performed using the color average value of the adjacent areas of the second target preset area to correct the color average value of the second target preset area; The second color average value is recalculated according to the corrected color average value, and the white balance parameter is calculated according to the re-obtained second color average value.

4. The 360-degree image adjustment method according to claim 1, characterized in that: The step of stitching the plurality of third images to form a final presented image comprises: When the third image is distorted, a first correction is performed according to the internal parameters of the camera, wherein the internal parameters include the focal length and the optical center position, and a second correction is performed on the third image after the first correction according to the relative position relationship between the cameras to obtain a fourth image; Detecting feature points in the fourth image and calculating descriptors of the feature points; Feature points in the plurality of fourth images are matched according to the descriptors to determine a relative position relationship between different fourth images, and the plurality of fourth images are spliced ​​according to the relative position relationship to obtain a final presented image.

5. The 360-degree image adjustment method according to claim 1, characterized in that: The method further comprises: Preprocessing the second image, and performing obstacle detection on the preprocessed second image according to a preset deep learning model, and outputting the category, position and confidence of the obstacle; Marking the dynamic obstacles in the image according to the positions and confidence levels of the obstacles; When the confidence level is higher than a threshold, a shadow removal algorithm is determined according to the shape, size and environment of the marked obstacle area, and shadow processing is performed on the obstacle area according to the shadow removal algorithm.

6. The 360-degree image adjustment method according to claim 1, characterized in that: The method further comprises: Analyzing the motion vector of each pixel in the second image to identify a dynamic obstacle, and recording the motion trajectory and speed of the dynamic obstacle; The future position of the dynamic obstacle is predicted according to the motion trajectory and speed, and the dynamic obstacle is marked in advance in subsequent images according to the future position so as to perform continuous shadow processing.

7. A 360-degree image adjustment system, characterized in that: It includes a detection module, an adjustment module, a calculation module and a display module, among which: a detection module configured to detect current light conditions through a light sensor or image data of a first image acquired by an analysis camera, and calculate target exposure parameters according to the current light conditions, wherein the target exposure parameters include shutter speed, aperture size, and sensitivity; an adjustment module, configured to adjust the camera according to the target exposure parameter, and acquire a second image through the adjusted camera; a calculation module configured to calculate color averages of different areas in the second image, and calculate white balance parameters according to the color averages, wherein the white balance parameters include color temperature, hue and saturation; a display module configured to perform color correction on the second image according to the white balance parameter to obtain a third image, and to splice a plurality of the third images to form a final displayed image, The step of calculating the color average values ​​of different regions in the second image and calculating the white balance parameter according to the color average values ​​further includes: Dividing the second image into a plurality of preset areas, calculating a first color average value of the three primary colors of red, green and blue in a first target preset area, wherein the first target preset area is any one of the plurality of preset areas; Determining a target weight according to the position and size of the first target preset area, and performing weighted averaging on the color averages of the plurality of preset areas based on the target weight and the first color average to obtain a second color average of the second image; The white balance parameter is determined according to the second color average value in combination with a preset mapping relationship between color and color temperature, hue, and saturation.

8. A controller, characterized in that: It includes a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are both used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the controller executes the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed by a computer device, the method according to any one of claims 1 to 6 is executed.

Images