Panoramic image brightness abnormal area detection and adjustment method

Abstract

The invention provides a panoramic image brightness abnormal region detection and adjustment method, which comprises the following steps: acquiring a panoramic image, extracting global brightness distribution characteristics and local brightness distribution characteristics of the panoramic image, and generating a brightness distribution diagram; identifying a high-brightness area in the image according to the brightness distribution diagram; according to the matching condition of the brightness value of the high-brightness area and the light source characteristic data, whether a brightness abnormal area exists or not is judged; according to the global features and the local features, identifying whether multiple light source superposition exists in the image, and generating a light source superposition area graph; according to whether the brightness distribution in the light source superposition area map accords with natural illumination characteristics, judging whether the light source superposition area map is marked as a brightness abnormal area; and according to the marking result of the brightness abnormal region, generating a brightness adjustment decision diagram for brightness adjustment of subsequent panoramic image synthesis, and ensuring that the brightness adjustment result is consistent with the natural illumination characteristic of the scene.

Description

Technical Field

[0001] The present invention relates to the field of lighting technology, and in particular to a method for detecting and adjusting abnormal brightness areas of panoramic images. Background Art

[0002] In the process of panoramic image synthesis, the core problem of the brightness discrimination mechanism is how to accurately distinguish between abnormal brightness areas and normal brightness areas in the image, especially under complex lighting conditions.

[0003] When there is a large area of ​​light source (such as a window) in the image, the area directly illuminated by the light source will form a high brightness area, while the brightness characteristics of the scene itself may be manifested as a uniform or gradual brightness distribution. In addition, when there are multiple light sources in the image, such as the superposition of natural light and artificial light, this will make the brightness distribution pattern more complex.

[0004] Therefore, how to accurately identify and adjust these abnormal brightness areas while maintaining the natural lighting effect of the image is an urgent problem to be solved. Summary of the invention

[0005] The purpose of the present invention is to address the above-mentioned problems and to propose a method for detecting and adjusting abnormal brightness areas in panoramic images. The method analyzes the global and local brightness distribution characteristics of panoramic images, combines time and seasonal factors, identifies high-brightness areas and determines whether they are directly illuminated by light sources or have natural brightness characteristics. The method can effectively identify and adjust abnormal brightness areas in panoramic images, improve image quality and realism, and is suitable for image processing in the fields of panoramic photography, virtual reality, etc.

[0006] The technical solution of the present invention is: The present invention provides a method for detecting and adjusting abnormal brightness areas of a panoramic image, comprising: A panoramic image is acquired, and global and local brightness distribution features of the panoramic image are extracted to generate a brightness distribution map; based on the brightness distribution map, a high-brightness area in the image is identified, and its spatial position and brightness value range are marked; in combination with time factors and seasonal factors, the light source intensity, angle and color temperature information of the current scene are acquired from a preset light source characteristic database; based on the matching between the brightness value of the high-brightness area and the light source characteristic data, it is determined whether there is a brightness abnormality area; based on the global and local features, it is determined whether there is a situation where multiple light sources are superimposed in the image, and a light source superposition area map is generated; based on whether the brightness distribution in the light source superposition area map conforms to the natural lighting characteristics, it is determined whether it is marked as a brightness abnormality area; based on the marking result of the brightness abnormality area, a brightness adjustment decision map is generated for brightness adjustment of subsequent panoramic image synthesis; a convolutional neural network is used to optimize the brightness adjustment decision map to ensure that the brightness adjustment result is consistent with the natural lighting characteristics of the scene.

[0007] Furthermore, the extracting of global brightness distribution features and local brightness distribution features of the panoramic image to generate a brightness distribution map includes: extracting the global brightness distribution features of the panoramic image by a histogram equalization algorithm to obtain global brightness feature data; extracting the local brightness distribution features of the panoramic image by a local binary pattern algorithm to obtain local brightness feature data; generating a brightness distribution feature matrix based on the global brightness feature data and the local brightness feature data; inputting the brightness distribution feature matrix into a convolutional neural network for feature fusion to obtain fused brightness features; and generating a brightness distribution map based on the fused brightness features using an image generation algorithm.

[0008] Furthermore, the method of identifying high-brightness areas in an image based on the brightness distribution map and marking their spatial positions and brightness value ranges includes: converting the brightness distribution map into a brightness map using a grayscale processing algorithm; extracting highlight areas from the brightness map using a preset brightness threshold segmentation algorithm; performing morphological operations on the highlight areas to remove noise areas and obtain accurate highlight area contours; calculating the spatial coordinates of each highlight area and recording its position value based on the highlight area contours; extracting the brightness value range of each highlight area and counting the minimum and maximum values ​​of the interval values; associating the position value with the interval value to generate highlight area marking information; and grouping the highlight areas using a clustering algorithm to determine their distribution characteristics and output a final marking result.

[0009] Furthermore, the method of obtaining the light source intensity, angle and color temperature information of the current scene from a preset light source characteristic database includes: determining the time and season factors in the environmental parameters based on the current time and season information; screening a light source data set that meets the conditions from the preset light source characteristic database based on the time and season factors; using a lighting condition analysis algorithm to process the screened light source data set to extract the light source intensity, angle and color temperature characteristics; if there are multiple matching items in the light source characteristics, calling a weighted average algorithm to calculate the average value of the light source intensity, determine the median value of the light source angle, and obtain the mode value of the light source color temperature; using an environmental parameter matching algorithm to compare the calculated light source characteristics with the current scene to determine whether the lighting conditions match; if the lighting conditions match, outputting the final results of the light source intensity, angle and color temperature.

[0010] Furthermore, the determination of whether there is an abnormal brightness area is based on the matching of the brightness value of the high-brightness area with the light source characteristic data, including: if the brightness value of the high-brightness area matches the light source characteristic data, determining that the area is a direct illumination area of ​​the light source and marking it as an abnormal brightness area; if the brightness value of the high-brightness area does not match the light source characteristic data, further analyzing its brightness distribution pattern to determine whether it is the natural brightness characteristic of the scene.

[0011] Furthermore, the method of identifying whether there are multiple light sources superimposed in the image based on the global features and the local features, and generating a light source superposition area map, includes: judging the number of light sources based on the global features and the local features, and determining a superposition state if the number of light sources is greater than 1; dividing the superposition state into regions using a preset threshold to obtain light source regions and superposition regions; performing feature matching on the light source regions and the superposition regions through an identifier to determine a superposition area map; and calculating area values ​​based on the area map to generate a light source superposition area map.

[0012] Furthermore, the brightness adjustment decision graph is optimized using a convolutional neural network to ensure that the brightness adjustment result is consistent with the natural lighting characteristics of the scene, including: using a preset convolutional neural network to process the brightness adjustment decision graph to generate an optimized decision graph; fusing the optimized decision graph with the scene graph to generate an adjusted scene graph; judging the consistency between the adjusted scene graph and the natural lighting characteristic value, and re-optimizing the decision graph if inconsistent.

[0013] The technical solution of the present invention includes the following beneficial effects: The present invention discloses a method for detecting and adjusting abnormal brightness areas in panoramic images. The method analyzes the global and local brightness distribution characteristics of panoramic images, combines time, seasonal factors and a preset light source characteristic database, and accurately identifies and marks high-brightness areas.

[0014] The present invention innovatively introduces light source superposition analysis, and effectively distinguishes normal brightness areas from abnormal areas by comparing brightness distribution with natural lighting characteristics. For the identified abnormal brightness areas, the present invention generates a brightness adjustment decision graph and optimizes it using a convolutional neural network to ensure that the adjustment result is consistent with the natural lighting characteristics of the scene.

[0015] This method can not only accurately detect brightness anomalies in panoramic images, but also intelligently adjust the brightness to improve the visual quality and realism of panoramic images, providing strong support for panoramic image synthesis and processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1 The present invention provides a method for detecting and adjusting abnormal brightness areas of a panoramic image, which specifically comprises the following steps: S101, acquiring a target panoramic image, extracting global brightness distribution features and local brightness distribution features of the image, and generating a brightness distribution map.

[0019] The panoramic image is acquired by an image acquisition device, and the panoramic image is preprocessed to obtain a high-definition panoramic image. The global brightness distribution characteristics of the high-definition panoramic image are extracted by a histogram equalization algorithm to obtain global brightness feature data. The local brightness distribution characteristics of the high-definition panoramic image are extracted by a local binary pattern algorithm to obtain local brightness feature data. A brightness distribution feature matrix is ​​generated based on the global brightness feature data and the local brightness feature data. The brightness distribution feature matrix is ​​input into a convolutional neural network for feature fusion to obtain a fused brightness feature. Based on the fused brightness feature, a brightness distribution map is generated by an image generation algorithm. The generated brightness distribution map is stored in a specified database.

[0020] S102: Identify high brightness areas in the image according to the brightness distribution map, and mark their spatial positions and brightness value ranges.

[0021] Obtain the brightness distribution map of the image, and use the grayscale processing algorithm to convert the image into a brightness map. Extract the highlight area from the brightness map through the preset brightness threshold segmentation algorithm. Perform morphological operations on the highlight area, remove the noise area, and obtain an accurate highlight area outline. According to the highlight area outline, calculate the spatial coordinates of each highlight area and record its position value. For each highlight area, extract its brightness value range, and count the minimum and maximum values ​​of the interval value. Associate the position value with the interval value to generate highlight area annotation information. Use a clustering algorithm to group the highlight areas, determine their distribution characteristics, and output the final annotation results.

[0022] S103 . Acquire the light source intensity, angle and color temperature information of the current scene from a preset light source characteristic database in combination with time factors and season factors.

[0023] Get the current time and season information, and determine the time and season factors in the environmental parameters. According to the time and season factors, select the light source data set that meets the conditions from the preset light source characteristic database. Use the lighting condition analysis algorithm to process the screened light source data set and extract the light source intensity, angle and color temperature characteristics. Light source characteristics refer to light source intensity, angle and color temperature characteristics; if there are multiple matching items for light source intensity, calculate the average value of light source intensity; if there are multiple matching items for light source angle, calculate the median value of light source angle; if there are multiple matching items for color temperature characteristics, obtain the mode value of light source color temperature. Through the environmental parameter matching algorithm, compare the calculated light source characteristics with the current scene to determine whether the lighting conditions match. If the lighting conditions match, output the final results of light source intensity, angle and color temperature; if not, re-screen the light source data set and repeat the feature extraction process. Store the final light source intensity, angle and color temperature information in the scene lighting database to complete the light source characteristic acquisition process of the current scene. 0024.S104. If the brightness value of the high-brightness area matches the light source characteristic data, the area is judged to be the area directly illuminated by the light source and marked as an abnormal brightness area; if the brightness value of the high-brightness area does not match the light source characteristic data, its brightness distribution pattern is further analyzed to determine whether it is the natural brightness characteristic of the scene.

[0024] .Calculate the brightness range of the light source direct area based on the light source characteristic data. If the brightness value of the high brightness area is within the range of the light source direct area, it is judged as the light source direct area. Mark the light source direct area as the brightness abnormal area. Use the region segmentation algorithm to divide the boundaries of the brightness abnormal area. Use image processing technology to remove the interference information of the brightness abnormal area. Generate a distribution map of the brightness abnormal area and store the processing results.

[0025] If the brightness value does not match the light source characteristic data, extract the brightness distribution pattern of the high brightness area. Use a clustering algorithm to classify the brightness distribution pattern and obtain the brightness distribution category. According to the brightness distribution category, determine whether it meets the natural brightness characteristics of the scene. If it meets the natural brightness characteristics of the scene, determine that the area is a natural high brightness area. If it does not meet the natural brightness characteristics of the scene, use an anomaly detection algorithm to identify abnormal brightness areas. Generate brightness correction parameters based on the distribution range of abnormal brightness areas.

[0026] In case the brightness value does not match the light source characteristics, the brightness distribution pattern of the high brightness area is extracted.

[0027] Specifically, the brightness distribution is obtained through image data, and the gray value parameters of high-brightness pixels are counted to generate a histogram, which can reflect the characteristics of brightness concentration or dispersion; the K-means clustering algorithm is used to divide the brightness values ​​into low-brightness, medium-brightness and high-brightness clusters according to the grayscale range; The natural brightness of indoor scenes should present a gradual characteristic rather than abrupt bright patches; if the clustering results show that the low-brightness, medium-brightness and high-brightness clusters present a gradual characteristic, and the brightness changes smoothly and conforms to the characteristics of ambient light distribution, then it is a natural brightness scene; otherwise, further identification is performed to determine whether it is a situation where multiple light sources are superimposed.

[0028] S105. Identify whether there is a superposition of multiple light sources in the image based on the global features and the local features, and generate a light source superposition area map.

[0029] Based on the global features and local features, feature points and feature values ​​are obtained; the number of light sources is determined based on the feature values. If the number of light sources is greater than 1, it is determined to be a superposition state. The superposition state is divided into regions using a preset threshold to obtain light source areas and superposition areas. The light source area and superposition area are feature matched by the identifier to determine the superposition area map. The area value is calculated based on the area map to generate a light source superposition area map. The light source superposition area map is compared with the image set to verify the accuracy of the area map. The light source superposition area map is output to complete the light source superposition identification and area map generation.

[0030] Global features reflect the overall brightness distribution of the image, such as brightness mean and contrast; local features focus on the texture or edge information of a specific area.

[0031] For example, in an indoor image, the global brightness mean is 150, and the grayscale values ​​of the feature points in the local highlight area are concentrated between 200 and 240.

[0032] If the eigenvalue shows two obvious bright peaks, corresponding to grayscale 234 and 240, it is inferred that there are two light sources; Set the grayscale threshold to 200. Areas above this value are marked as light source areas, and areas between 180-200 may be overlapping areas. Through threshold division, the main light area and the overlapping area are clearly distinguished. The threshold can be dynamically adjusted according to the scene lighting to ensure accurate division.

[0033] Furthermore, the identifier performs feature matching on the light source area and the superposition area to generate a superposition area map. The identifier compares the feature points of the light source area with the preset light source model based on the template. The feature points of the main light area present a circular bright spot, and the matching degree with the model reaches 90%, confirming that it is the light source area. The feature points of the superposition area show a mixed brightness mode, and are marked as the superposition area after matching.

[0034] The regional value is calculated according to the overlapping area, and is obtained by counting the pixel proportion and grayscale mean of each area. For example, the light source area accounts for 10% of the image, and the grayscale mean is 215; the overlapping area accounts for 5%, and the mean is 190. The regional map is presented in the form of a heat map, with the highlighted area marked in red and the overlapping area in yellow, which is convenient for intuitive analysis. The output regional map clearly marks the main light area, auxiliary light area and its overlapping area, and the regional boundaries are clear at a glance. This method ensures the accurate identification of light source distribution through multi-level feature analysis, providing a reliable basis for subsequent image processing.

[0035] S106: If the brightness distribution in the light source superposition area map conforms to the natural lighting characteristics, it is not marked as a brightness abnormal area; otherwise, it is marked as a brightness abnormal area.

[0036] Obtain the brightness distribution data of the light source superposition area, use the preset natural light characteristic model to analyze the brightness distribution data, and determine the degree of conformity between the brightness distribution data and the natural light characteristic model. If the brightness distribution data conforms to the natural light characteristic model, the area is determined to be a normal brightness area. If the brightness distribution data does not conform to the natural light characteristic model, the area is determined to be an abnormal brightness area. Generate a brightness abnormal area marking map based on the distribution characteristics of the abnormal brightness area. Smooth the abnormal brightness area marking map through the image processing algorithm to obtain the final abnormal brightness area marking result.

[0037] S107: Generate a brightness adjustment decision map according to the marking result of the abnormal brightness area, which is used for brightness adjustment of subsequent panoramic image synthesis.

[0038] Obtain the brightness value of the panoramic image, divide the area according to the brightness value, and determine the marking points of the abnormal area; generate a regional map based on the location information of the marking points; use the data of the regional map and the preset brightness adjustment value to generate a decision map. According to the instructions of the decision map, adjust the brightness value of the panoramic image to obtain a composite image. By comparing the composite image with the original image, judge the brightness adjustment effect. If the brightness adjustment effect does not meet the preset standard, regenerate the decision map and adjust the brightness again. Finally, obtain a panoramic composite image that meets the brightness standard.

[0039] S108. Use a convolutional neural network to optimize the brightness adjustment decision graph to ensure that the brightness adjustment result is consistent with the natural lighting characteristics of the scene.

[0040] Get the scene graph to be processed and extract the brightness graph in the scene graph. Use the preset convolutional neural network to process the brightness graph and generate an initial decision graph. Optimize the initial decision graph according to the natural light characteristic value to obtain an optimized decision graph. Fuse the optimized decision graph with the scene graph to generate an adjusted scene graph. Determine the consistency between the adjusted scene graph and the natural light characteristic value. If not, re-optimize the decision graph. Improve the accuracy of brightness adjustment by adjusting the parameters of the convolutional neural network. Finally, output a scene graph consistent with the natural light characteristics.

[0041] The detection method of the present invention determines whether it is a directly illuminated area by the light source and marks it as an abnormal brightness area by analyzing the degree of match between the brightness value of the high brightness area in the image and the light source characteristic data. For the unmatched high brightness area, it is further determined whether it is the natural brightness characteristic of the scene; if it does not meet the requirements, it identifies whether there are multiple light sources superimposed in the image and generates a light source superposition area map; it can effectively distinguish between normal lighting and abnormal lighting, accurately identify abnormal brightness areas in complex lighting environments, and provide a comprehensive and accurate solution for light source abnormality detection and lighting environment analysis.

[0042] It is obvious to those skilled in the art that the present application is not limited to the details of the exemplary embodiments described above, and that the present application can be implemented in other specific forms without departing from the spirit or essential features of the present application. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present application.

Claims

1. A method for detecting and adjusting abnormal brightness areas in a panoramic image, characterized in that: include: Acquire a panoramic image, extract global brightness distribution features and local brightness distribution features of the panoramic image, and generate a brightness distribution map; According to the brightness distribution map, identify the high brightness area in the image and mark its spatial position and brightness value range; Combined with time and season factors, the light intensity, angle and color temperature information of the current scene are obtained from the preset light source characteristic database; According to the matching condition between the brightness value of the high brightness area and the light source characteristic data, determining whether there is an abnormal brightness area; Based on global features and local features, identify whether there are multiple light sources superimposed in the image and generate a light source superposition area map; Determine whether to mark it as an abnormal brightness area according to whether the brightness distribution in the light source superposition area map conforms to the natural lighting characteristics; Generating a brightness adjustment decision map according to the marking result of the abnormal brightness area for brightness adjustment of subsequent panoramic image synthesis; The brightness adjustment decision graph is optimized using a convolutional neural network to ensure that the brightness adjustment result is consistent with the natural lighting characteristics of the scene; The method of identifying whether there are multiple light sources superimposed in the image based on the global features and the local features, and generating a light source superimposed area map, includes: According to the global features and the local features, the number of light sources is determined, and if the number of light sources is greater than 1, it is determined to be a superposition state; Using a preset threshold value to divide the superposition state into regions to obtain a light source region and a superposition region; Performing feature matching between the light source area and the superposition area through an identifier to determine a superposition area map; The area value is calculated according to the area map to generate a light source superposition area map.

2. The method according to claim 1, characterized in that The step of extracting global brightness distribution features and local brightness distribution features of the panoramic image to generate a brightness distribution map includes: Extracting global brightness distribution characteristics of the panoramic image by a histogram equalization algorithm to obtain global brightness feature data; Extracting local brightness distribution characteristics of the panoramic image using a local binary pattern algorithm to obtain local brightness feature data; Generate a brightness distribution feature matrix according to the global brightness feature data and the local brightness feature data; Inputting the brightness distribution feature matrix into a convolutional neural network for feature fusion to obtain fused brightness features; According to the fused brightness features, a brightness distribution map is generated using an image generation algorithm.

3. The method according to claim 1, characterized in that The step of identifying a high brightness area in the image according to the brightness distribution map and marking its spatial position and brightness value range includes: Converting the brightness distribution map into a brightness map using a grayscale processing algorithm; Extracting a highlight area from the brightness image by using a preset brightness threshold segmentation algorithm; Performing morphological operations on the highlight area to remove noise areas and obtain an accurate outline of the highlight area; According to the outline of the highlight area, calculate the spatial coordinates of each highlight area and record its position value; For each highlight area, extract its brightness value range and calculate the minimum and maximum values ​​of the interval value; Associating the position value with the interval value to generate highlight area marking information; A clustering algorithm is used to group the highlighted areas, determine their distribution characteristics, and output the final labeling results.

4. The method according to claim 1, characterized in that The step of obtaining the light source intensity, angle and color temperature information of the current scene from a preset light source characteristic database includes: Determine the time and season factors in environmental parameters based on current time and season information; According to the time and season factors, a light source data set that meets the conditions is selected from a preset light source characteristic database; Using the illumination condition analysis algorithm, the screened light source data set is processed to extract the light source intensity, angle and color temperature characteristics; If there are multiple matching items for the light source characteristics, the weighted average algorithm is called to calculate the average value of the light source intensity, determine the median value of the light source angle, and obtain the mode value of the light source color temperature; Through the environmental parameter matching algorithm, the calculated light source characteristics are compared with the current scene to determine whether the lighting conditions match; If the lighting conditions match, the final results of light source intensity, angle and color temperature are output.

5. The method according to claim 1, characterized in that The determining whether there is an abnormal brightness area according to the matching condition between the brightness value of the high brightness area and the light source characteristic data includes: If the brightness value of the high brightness area matches the light source characteristic data, the area is determined to be a direct illumination area of ​​the light source and marked as an abnormal brightness area; If the brightness value of the high brightness area does not match the light source characteristic data, the brightness distribution pattern is further analyzed to determine whether it is the natural brightness characteristic of the scene.

6. The method according to claim 1, characterized in that The use of a convolutional neural network to optimize the brightness adjustment decision graph to ensure that the brightness adjustment result is consistent with the natural lighting characteristics of the scene includes: Using a preset convolutional neural network to process the brightness adjustment decision graph to generate an optimized decision graph; Fusing the optimized decision graph with the scene graph to generate an adjusted scene graph; The consistency between the adjusted scene graph and the natural lighting characteristic value is determined, and if not, the decision graph is re-optimized.

Images