A method and device for roughly distinguishing cloud cover based on satellite image thumbnails

Abstract

The present invention provides a method and device for roughly distinguishing cloud cover based on satellite image thumbnails, and relates to the technical field of image processing. The method comprises: after image preprocessing is performed on a to-be-processed satellite image, a multispectral image and a full-color image in the to-be-processed satellite image are fused to obtain a satellite fused image; scaling the satellite fused image by a bicubic interpolation method is performed to obtain a satellite thumbnail; image segmentation is performed on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, and then cloud cover features of each pixel point in the image block are extracted, wherein the cloud cover features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer; based on the cloud cover features of each pixel point in the image block, cloud cover in each image block is evaluated to determine a rough cloud cover distinguishing result of the satellite image thumbnail.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and device for roughly distinguishing cloud cover based on satellite image thumbnails. Background Art

[0002] Clouds are inevitable in optical remote sensing images. The amount of cloud affects the use of image products. Cloud cover affects the accurate acquisition of ground object information and causes a large amount of useless remote sensing data transmission and processing, wasting storage and computing resources. Therefore, cloud detection is an indispensable part of satellite data quality evaluation.

[0003] In the related technologies, during the cloud cover detection process, the remote sensing image needs to restore the reflectivity through radiation correction, atmospheric correction, etc., and the radiation coefficient of each satellite is different. The radiation coefficient also needs to be updated regularly, which limits the versatility of the algorithm and increases the complexity of the system. Therefore, how to efficiently identify the cloud cover of satellite remote sensing images has become a problem that needs to be solved urgently in the industry. Summary of the invention

[0004] The present invention provides a method and device for roughly distinguishing cloud cover based on satellite image thumbnails, which are used to solve the problem of how to efficiently distinguish cloud cover of satellite remote sensing images in the prior art.

[0005] The present invention provides a method for roughly distinguishing cloud cover based on satellite image thumbnails, comprising:

[0006] After the satellite image to be processed is preprocessed, the multispectral image and the panchromatic image in the satellite image to be processed are fused to obtain a satellite fused image;

[0007] Scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0008] Performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0009] Based on the cloud cover characteristics of each pixel in the image block, cloud cover in each image block is evaluated to determine a rough cloud cover determination result of the satellite image thumbnail.

[0010] According to a cloud amount rough discrimination method based on satellite image thumbnails provided by the present invention, the image segmentation method of N image blocks includes:

[0011] By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained;

[0012] Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value;

[0013] When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained;

[0014] According to the pixel areas divided by the watershed pixel points, N image blocks are obtained after the image is segmented.

[0015] According to a rough cloud amount discrimination method based on satellite image thumbnails provided by the present invention, based on the cloud amount characteristics of each pixel point in the image block, cloud amount in each image block is evaluated to determine the rough cloud amount discrimination result of the satellite image thumbnail, including:

[0016] For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard;

[0017] Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block;

[0018] A rough cloud amount determination result of the satellite image thumbnail is determined according to the cloud coverage area in each image block.

[0019] According to a cloud amount rough discrimination method based on satellite image thumbnails provided by the present invention, a cloud amount rough discrimination result of the satellite image thumbnail is determined according to the cloud coverage area in each image block, including:

[0020] Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks;

[0021] The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail.

[0022] According to a cloud amount rough discrimination method based on satellite image thumbnails provided by the present invention, after the satellite image to be processed is subjected to image preprocessing, before the step of fusing the multispectral image and the panchromatic image in the satellite image to be processed to obtain the satellite fused image, the method further comprises:

[0023] Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade;

[0024] The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

[0025] According to a cloud amount rough discrimination method based on satellite image thumbnails provided by the present invention, after the satellite image to be processed is subjected to image preprocessing, the multispectral image and the panchromatic image in the satellite image to be processed are subjected to image fusion to obtain a satellite fusion image, comprising:

[0026] When the multispectral image and the panchromatic image match in spatial position, perform principal component transformation on the multispectral image, calculate the covariance matrix or correlation matrix of the multispectral image, and solve the eigenvalues ​​and corresponding eigenvectors of the matrix;

[0027] The eigenvectors corresponding to the first M eigenvalues ​​in the eigenvalue sorting are histogram matched and replaced with the full-color image to obtain a satellite fusion image.

[0028] The present invention also provides a device for roughly distinguishing cloud cover based on satellite image thumbnails, comprising:

[0029] A fusion module is used to fuse the multispectral image and the panchromatic image in the satellite image to be processed after image preprocessing to obtain a satellite fusion image;

[0030] A scaling module, used for scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0031] An extraction module is used to perform image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, and then extract cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0032] The evaluation module is used to evaluate the cloud amount in each image block based on the cloud amount characteristics of each pixel point in the image block, so as to determine the rough judgment result of the cloud amount of the satellite image thumbnail.

[0033] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0034] By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained;

[0035] Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value;

[0036] When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained;

[0037] According to the pixel areas divided by the watershed pixel points, N image blocks are obtained after the image is segmented.

[0038] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0039] For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard;

[0040] Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block;

[0041] A rough cloud amount determination result of the satellite image thumbnail is determined according to the cloud coverage area in each image block.

[0042] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0043] Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks;

[0044] The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail.

[0045] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0046] Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade;

[0047] The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

[0048] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0049] When the multispectral image and the panchromatic image match in spatial position, perform principal component transformation on the multispectral image, calculate the covariance matrix or correlation matrix of the multispectral image, and solve the eigenvalues ​​and corresponding eigenvectors of the matrix;

[0050] The eigenvectors corresponding to the first M eigenvalues ​​in the eigenvalue sorting are histogram matched and replaced with the full-color image to obtain a satellite fusion image.

[0051] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a rough cloud amount determination method based on satellite image thumbnails as described above is implemented.

[0052] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for roughly distinguishing cloud cover based on satellite image thumbnails as described in any one of the above is implemented.

[0053] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for roughly distinguishing cloud cover based on satellite image thumbnails.

[0054] The cloud amount rough discrimination method and device based on satellite image thumbnails provided by the present invention can effectively improve image quality, reduce image distortion and pixelation effects through image fusion scaling processing, and can effectively segment thumbnails through image watershed algorithm, decomposing the image into N image blocks. Different areas in the image are identified, and based on the extracted cloud amount features, the image watershed algorithm can also be used to evaluate the cloud amount of each image block. Finally, the cloud amount evaluation results of each image block are combined to determine the cloud amount rough discrimination result of the entire satellite image thumbnail, without excessively considering the influence of parameters such as the radiation coefficient, and the efficiency and accuracy of cloud amount evaluation are improved through automated methods, which is of great significance for the preprocessing and screening of remote sensing data. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0056] Figure 1 A schematic diagram of the flow of a method for roughly distinguishing cloud cover based on satellite image thumbnails provided in an embodiment of the present application;

[0057] Figure 2 A schematic diagram of the structure of a device for roughly distinguishing cloud cover based on satellite image thumbnails provided in an embodiment of the present application;

[0058] Figure 3 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Figure 1 A schematic diagram of a rough cloud amount determination method based on satellite image thumbnails provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, including:

[0061] Step 110, after the satellite image to be processed is preprocessed, the multispectral image and the panchromatic image in the satellite image to be processed are fused to obtain a satellite fused image;

[0062] In the embodiment of the present application, the satellite image to be processed may specifically refer to a remote sensing image taken by a satellite. The image preprocessing in the embodiment of the present application is intended to improve the image quality and provide clearer and more accurate data for subsequent processing steps. The image preprocessing may specifically include:

[0063] Denoising: Use filters to remove random noise in an image and improve the signal-to-noise ratio of the image. For example, filtering can be performed on satellite images to remove noise and make the image clearer.

[0064] Distortion Correction: Adjusts the geometry of processed satellite imagery to ensure that objects in the image have the correct scale and proportion.

[0065] Noise Reduction: Further reduce the noise in the satellite image to be processed and retain the detailed features of the image.

[0066] In the embodiment of the present application, image fusion is the combination of multispectral images and panchromatic images to improve the spatial resolution and spectral information of the image. Therefore, the satellite fusion image in the present application retains the high spatial resolution of the panchromatic image and contains the rich spectral information of the multispectral image, which is very helpful for improving the interpretability and analysis accuracy of the image.

[0067] Step 120, scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0068] In the embodiment of the present application, the bicubic interpolation method is to establish a polynomial surface around the pixels of the original image to predict the values ​​of new pixels, thereby generating a high-resolution or low-resolution image.

[0069] Specifically, the scaling process may include:

[0070] Step 1: Define the coordinate system of the satellite fusion image and the coordinate system of the target image, which are (x, y) and (u, v) respectively.

[0071] Step 2: For each pixel in the target image, determine its corresponding coordinates (x, y) in the original image.

[0072] Step 3: Select 16 pixels in the original image and use the coordinates (x, y) and pixel values ​​f(x, y) of these 16 pixels to calculate coefficients a0 to a9.

[0073] Step 4: Substitute the calculated coefficients into the formula f(x,y) to find the pixel value of the pixel in the target image.

[0074] Step 5: Repeat steps 3 and 4 to find the pixel values ​​in all target images.

[0075] In an embodiment of the present application, the interpolation algorithm may perform image preprocessing in advance, such as noise removal, smoothing, etc., to ensure the accuracy of interpolation.

[0076] f(x,y)=a0+a1x+a2y+a3xy+a4x^2+a5y^2+a6x^2y+a7xy^2+a8x^3+a9y^3

[0077] Where f(x,y) is the pixel value of the image, and a0 to a9 are the coefficients to be determined.

[0078] In this application, after the pixel values ​​in the target image are obtained, the satellite thumbnail can be obtained.

[0079] Step 130, performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, and then extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0080] In an embodiment of the present application, the image watershed algorithm is a mathematical morphological image segmentation technology based on image grayscale values, which can be used to segment satellite thumbnails into multiple image blocks, which can represent different land features, such as clouds, land, water bodies, etc.

[0081] In the embodiment of the present application, after the image is segmented and N image blocks are obtained, feature extraction is performed on the pixels in each image block to identify and quantify the features of the cloud. The features may include:

[0082] Gray value features: calculate the gray histogram, average gray value, gray co-occurrence matrix, etc. of the image block.

[0083] Morphological features: Analyze the morphological properties of image blocks, such as shape, size, convexity, etc.

[0084] Color space features: Analyze the color distribution and color histogram of image blocks in different color spaces (such as RGB, HSV).

[0085] Texture features: Calculate texture descriptors such as contrast, uniformity, and directionality of the gray-level co-occurrence matrix to describe the texture structure of the image block.

[0086] The extracted features are used for analysis to identify and quantify cloud cover. For example, clouds usually have high grayscale values ​​and specific texture features. Based on the analysis results, the pixels belonging to clouds are marked in preparation for cloud cover assessment.

[0087] Step 140 , based on the cloud cover features of each pixel in the image block, perform cloud cover assessment in each image block to determine a rough cloud cover determination result of the satellite image thumbnail.

[0088] In an embodiment of the present application, cloud amount features are extracted for pixel points in each image block, and these features may include gray value features, morphological features, color space features, and texture features.

[0089] The extracted cloud features are analyzed to identify pixels that may represent clouds. For example, clouds usually appear as areas of high grayscale values ​​in grayscale images and have specific characteristics in color space.

[0090] Using classification algorithms, such as support vector machines (SVM), random forests, or neural networks, pixels are classified as cloud pixels or non-cloud pixels based on the extracted features.

[0091] For each image block, the proportion or number of cloud pixels is calculated to quantify the degree of cloud cover in the image block. A threshold for cloud cover is set according to application requirements to determine whether an image block is considered to have high or low cloud cover.

[0092] According to the proportion or number of cloud pixels, the cloud cover of each image block is evaluated to determine its cloud cover level. The cloud cover evaluation results of all image blocks are combined to obtain the rough cloud cover judgment result of the entire satellite image thumbnail. According to the combined cloud cover evaluation results, the satellite image thumbnail is divided into different cloud cover levels, such as no cloud, few clouds, cloudy, etc.

[0093] In the embodiment of the present application, the image quality can be effectively improved, image distortion and pixelation effects can be reduced through the scaling process of image fusion, and the thumbnail can be effectively segmented through the image watershed algorithm to decompose the image into N image blocks. Different areas in the image are identified, and based on the extracted cloud cover features, the image watershed algorithm can also be used to evaluate the cloud cover of each image block. Finally, the cloud cover evaluation results of each image block are combined to determine the rough judgment result of the cloud cover of the entire satellite image thumbnail. There is no need to consider the influence of parameters such as the radiation coefficient too much. The efficiency and accuracy of cloud cover evaluation are improved through automated methods, which is of great significance for the preprocessing and screening of remote sensing data.

[0094] Optionally, the image segmentation method of the N image blocks includes:

[0095] By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained;

[0096] Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value;

[0097] When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained;

[0098] According to the pixel areas divided by the watershed pixel points, N image blocks are obtained after the image is segmented.

[0099] In the embodiment of the present application, an energy function is constructed according to the grayscale value of the satellite thumbnail. This function is related to the gradient of the image, that is, the rate of grayscale change. Then, the energy state of each pixel in the satellite thumbnail is calculated. The energy state can be expressed as the value of the pixel in the energy function.

[0100] Select P seed pixels in the satellite thumbnail randomly or according to some criteria. These seed pixels can be regarded as the source of water and are used to simulate the diffusion of water. Starting from the selected seed pixels, simulate the diffusion of water to the surrounding adjacent pixels with lower energy values. Diffusion follows the guidance of the energy function, and water flows from high energy to low energy. During the diffusion simulation, update the "water level" of each pixel, that is, the degree to which water covers the pixel. When "water" from different seed pixels meets, a watershed pixel is formed at the meeting point. These points represent the boundaries of different areas in the image. Identify and record all watershed pixels.

[0101] According to the watershed pixels, the image is divided into different regions. Each region is composed of pixels with similar energy states. Through the watershed algorithm, N image blocks are finally obtained after image segmentation, and each image block represents an independent region in the image. The segmentation results are post-processed, such as removing small, meaningless areas, or merging adjacent areas with similar attributes, to obtain N image blocks after image segmentation.

[0102] In the embodiment of the present application, the image watershed algorithm can be used to effectively segment the image, which is beneficial for subsequent cloud cover determination.

[0103] Optionally, based on the cloud cover feature of each pixel in the image block, cloud cover evaluation is performed in each image block to determine a rough cloud cover determination result of the satellite image thumbnail, including:

[0104] For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard;

[0105] Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block;

[0106] A rough cloud amount determination result of the satellite image thumbnail is determined according to the cloud coverage area in each image block.

[0107] In an embodiment of the present application, the preset cloud amount feature standard may specifically be a standard pre-set by a user that conforms to cloud pixel characteristics, and may specifically include a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard.

[0108] In an optional embodiment, the gray value feature standard is: clouds are usually brighter in grayscale images.

[0109] Morphological characteristics criteria: morphological characteristics of clouds, such as shape and size.

[0110] Color space feature standards: Cloud features in different color spaces (such as RGB, HSV).

[0111] Texture characteristic criteria: Texture properties of the cloud region, such as uniformity and directionality.

[0112] In an embodiment of the present application, for each image block, the pixels therein are checked, and the pixels that meet the cloud characteristics are identified according to a preset cloud feature standard.

[0113] The identified cloud pixels are traversed in the row and column directions until a non-cloud pixel is encountered or the image boundary is reached. Through the traversal process, the cloud coverage area in each image block is determined. This may involve marking consecutive cloud pixels to form connected areas of the cloud, and then determining the cloud coverage area in each image block based on the connected areas.

[0114] Optionally, determining a rough cloud amount determination result of the satellite image thumbnail according to the cloud coverage area in each image block includes:

[0115] Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks;

[0116] The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail.

[0117] In the embodiment of the present application, after determining each cloud coverage area, the non-cloud area introduced by the highlight objects can also be removed according to the cloud boundary shape.

[0118] More specifically, for each image block, the number of pixels or area percentage of cloud coverage is calculated. This is usually achieved by counting the number of cloud pixels and dividing it by the total number of pixels in the image block.

[0119] According to the proportion of cloud coverage area, the cloud cover of each image block is roughly judged. Thresholds can be set to classify cloud cover, for example: no cloud: cloud coverage ratio is lower than a certain threshold (such as 5%); few clouds: cloud coverage ratio is within a certain range (such as 5% to 25%); cloudy: cloud coverage ratio exceeds a certain threshold (such as more than 25%). According to the calculated cloud coverage ratio, a cloud cover level is assigned to each image block.

[0120] More specifically, the cloud amount rough identification results of all image blocks can also be combined to obtain the cloud amount situation of the entire satellite image thumbnail. For example, the cloud amount evaluation results of each image block are weighted averaged according to the area ratio of the image block in the thumbnail.

[0121] Based on the merged results, an overall cloud cover assessment is made for the entire satellite image thumbnail. This can be an overall cloud cover level, such as "overall few clouds" or "overall cloudy".

[0122] Satellite image thumbnails are rated for quality based on the overall cloud cover assessment, which helps determine whether the image is suitable for further analysis and applications.

[0123] In the embodiment of the present application, the cloud cover of the satellite image thumbnail can be effectively evaluated through the rough judgment result of the cloud cover of the satellite image thumbnail, providing an important basis for the screening, classification and subsequent analysis of remote sensing data.

[0124] Optionally, before the step of fusing the multispectral image and the panchromatic image in the satellite image to be processed to obtain the satellite fused image after image preprocessing, the method further comprises:

[0125] Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade;

[0126] The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

[0127] In an embodiment of the present application, the preset cloud amount assessment standard is used to clarify the cloud amount evaluation standard and cloud amount assessment level, and these standards may be determined based on the percentage of cloud amount, cloud type, cloud distribution, etc.

[0128] By screening with preset cloud cover assessment standards, a lot of cloudless data can be effectively avoided, thus avoiding invalid data analysis.

[0129] In an optional embodiment, the initial satellite image is subjected to absolute radiometric correction. The purpose of this step is to eliminate the error of the sensor itself and ensure that the obtained radiometric brightness value accurately reflects the real radiometric characteristics of the ground target.

[0130] The absolute radiation correction may specifically include: determining the radiation gain and bias parameters of the sensor.

[0131] Apply the radiation correction formula to convert the original digital value (DN value) into a radiation brightness value.

[0132] The correction formula is as follows:

[0133] L λ =k*DN+c

[0134] Among them, L λ is the radiation brightness value of band λ, k is the radiation gain, and c is the radiation bias.

[0135] In the embodiment of the present application, after the absolute radiation correction is completed, the corrected satellite images to be processed are obtained, and these images can now be used for subsequent image preprocessing and image fusion steps.

[0136] Optionally, after the satellite image to be processed is preprocessed, the multispectral image and the panchromatic image in the satellite image to be processed are fused to obtain a satellite fused image, including:

[0137] When the multispectral image and the panchromatic image match in spatial position, perform principal component transformation on the multispectral image, calculate the covariance matrix or correlation matrix of the multispectral image, and solve the eigenvalues ​​and corresponding eigenvectors of the matrix;

[0138] The eigenvectors corresponding to the first M eigenvalues ​​in the eigenvalue sorting are histogram matched and replaced with the full-color image to obtain a satellite fusion image.

[0139] In the embodiment of the present application, the spatial position matching of the multispectral image and the panchromatic image is used to ensure that the multispectral image and the panchromatic image are matched in spatial position, that is, they cover the same geographical area and the pixels correspond.

[0140] Perform principal component transformation on the multispectral image and calculate its covariance matrix or correlation matrix based on the data of the multispectral image. These matrices describe the linear relationship between different spectral bands. Solve the eigenvalues ​​and corresponding eigenvectors of the covariance matrix or correlation matrix. The eigenvalues ​​represent the contribution of each principal component, while the eigenvectors represent the projection direction of the data on these principal components.

[0141] More specifically, the obtained eigenvalues ​​can be sorted from large to small, the first M eigenvalues ​​are selected, and the corresponding first M eigenvectors are selected according to the sorted eigenvalues. These eigenvectors represent the most important change patterns in the data. The data of the panchromatic image is histogram matched with the selected eigenvectors. The statistical characteristics of the panchromatic image are adjusted to be consistent with the main change patterns of the multispectral image.

[0142] The corresponding feature vectors in the multispectral image are replaced with the matched panchromatic image data to generate a new fused image dataset, and the replaced feature vectors are recombined to generate a satellite fused image with high spatial resolution and rich spectral information.

[0143] In the embodiment of the present application, the spectral information of the multispectral image and the spatial resolution of the panchromatic image are effectively combined to obtain a satellite fusion image with better characteristics, which is beneficial to cloud cover discrimination.

[0144] The following is a description of the cloud amount coarse discrimination device based on satellite image thumbnails provided by the present invention. The cloud amount coarse discrimination device based on satellite image thumbnails described below and the cloud amount coarse discrimination method based on satellite image thumbnails described above can be referenced to each other.

[0145] Figure 2 The schematic diagram of the structure of the cloud amount rough determination device based on satellite image thumbnails provided in the embodiment of the present application is as follows: Figure 2 As shown, including:

[0146] The fusion module 210 is used to fuse the multispectral image and the panchromatic image in the satellite image to be processed after image preprocessing to obtain a satellite fusion image;

[0147] The scaling module 220 is used to scale the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0148] The extraction module 230 is used to perform image segmentation on the satellite thumbnail according to the image watershed algorithm, and after obtaining N image blocks, extract the cloud amount feature of each pixel in the image block, wherein the cloud amount feature includes at least one of the following: gray value feature, morphological feature, color space feature, texture feature, and N is a positive integer;

[0149] The evaluation module 240 is used to evaluate the cloud cover in each image block based on the cloud cover feature of each pixel in the image block, so as to determine the rough judgment result of the cloud cover of the satellite image thumbnail.

[0150] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0151] By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained;

[0152] Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value;

[0153] When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained;

[0154] According to the pixel areas divided by the watershed pixel points, N image blocks are obtained after the image is segmented.

[0155] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0156] For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard;

[0157] Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block;

[0158] A rough cloud amount determination result of the satellite image thumbnail is determined according to the cloud coverage area in each image block.

[0159] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0160] Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks;

[0161] The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail.

[0162] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0163] Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade;

[0164] The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

[0165] According to an embodiment of the present invention, a device for roughly distinguishing cloud cover based on satellite image thumbnails is provided, and the device is also used for:

[0166] When the multispectral image and the panchromatic image match in spatial position, perform principal component transformation on the multispectral image, calculate the covariance matrix or correlation matrix of the multispectral image, and solve the eigenvalues ​​and corresponding eigenvectors of the matrix;

[0167] The eigenvectors corresponding to the first M eigenvalues ​​in the eigenvalue sorting are histogram matched and replaced with the full-color image to obtain a satellite fusion image.

[0168] In the embodiment of the present application, the image quality can be effectively improved, image distortion and pixelation effects can be reduced through the scaling process of image fusion, and the thumbnail can be effectively segmented through the image watershed algorithm to decompose the image into N image blocks. Different areas in the image are identified, and based on the extracted cloud cover features, the image watershed algorithm can also be used to evaluate the cloud cover of each image block. Finally, the cloud cover evaluation results of each image block are combined to determine the rough judgment result of the cloud cover of the entire satellite image thumbnail. There is no need to consider the influence of parameters such as the radiation coefficient too much. The efficiency and accuracy of cloud cover evaluation are improved through automated methods, which is of great significance for the preprocessing and screening of remote sensing data.

[0169] Figure 3 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 3 As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330 and a communication bus 340, wherein the processor 310, the communication interface 320 and the memory 330 communicate with each other through the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute a rough cloud amount discrimination method based on satellite image thumbnails, the method comprising: after image preprocessing of the satellite image to be processed, performing image fusion on the multispectral image and the panchromatic image in the satellite image to be processed to obtain a satellite fusion image;

[0170] Scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0171] Performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0172] Based on the cloud cover characteristics of each pixel in the image block, cloud cover in each image block is evaluated to determine a rough cloud cover determination result of the satellite image thumbnail.

[0173] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0174] On the other hand, the present invention also provides a computer program product, the computer program product includes a computer program, the computer program can be stored in a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the cloud amount rough discrimination method based on satellite image thumbnails provided by the above methods, the method comprising: after image preprocessing of the satellite image to be processed, fusing the multispectral image and the panchromatic image in the satellite image to be processed to obtain a satellite fused image;

[0175] Scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0176] Performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0177] Based on the cloud cover characteristics of each pixel in the image block, cloud cover in each image block is evaluated to determine a rough cloud cover determination result of the satellite image thumbnail.

[0178] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the cloud amount rough discrimination method based on satellite image thumbnails provided by the above methods, the method comprising: after image preprocessing of the satellite image to be processed, fusing the multispectral image and the panchromatic image in the satellite image to be processed to obtain a satellite fused image;

[0179] Scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail;

[0180] Performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer;

[0181] Based on the cloud amount characteristics of each pixel in the image block, the cloud amount in each image block is evaluated to determine the rough cloud amount discrimination result of the satellite image thumbnail.

[0182] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0183] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0184] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rough cloud determination method based on satellite image thumbnails. It is characterized in that include: After the satellite image to be processed is preprocessed, the multispectral image and the panchromatic image in the satellite image to be processed are fused to obtain a satellite fused image; Scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail; Performing image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, extracting cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer; Based on the cloud amount characteristics of each pixel point in the image block, cloud amount evaluation is performed in each image block to determine a rough cloud amount discrimination result of the satellite image thumbnail; The image segmentation method of the N image blocks includes: By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained; Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value; When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained; Obtaining N image blocks after image segmentation according to the pixel areas divided by each of the watershed pixel points; Wherein, based on the cloud amount characteristics of each pixel point in the image block, cloud amount evaluation in each image block is performed to determine the rough cloud amount discrimination result of the satellite image thumbnail, including: For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard; Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block; Determining a rough cloud amount determination result of the satellite image thumbnail according to the cloud coverage area in each image block; Wherein, determining the rough cloud amount determination result of the satellite image thumbnail according to the cloud coverage area in each image block includes: Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks; The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail; Wherein, before the step of fusing the multispectral image and the panchromatic image in the satellite image to be processed after image preprocessing to obtain the satellite fused image, the method further comprises: Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade; The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

2. The cloud cover rough identification method based on satellite image thumbnails according to claim 1, It is characterized in that After the satellite image to be processed is preprocessed, the multispectral image and the panchromatic image in the satellite image to be processed are fused to obtain a satellite fused image, including: When the multispectral image and the panchromatic image match in spatial position, perform principal component transformation on the multispectral image, calculate the covariance matrix or correlation matrix of the multispectral image, and solve the eigenvalues ​​and corresponding eigenvectors of the matrix; The eigenvectors corresponding to the first M eigenvalues ​​in the eigenvalue sorting are histogram matched and replaced with the full-color image to obtain a satellite fusion image.

3. A rough cloud determination device based on satellite image thumbnails, It is characterized in that include: A fusion module is used to fuse the multispectral image and the panchromatic image in the satellite image to be processed after image preprocessing to obtain a satellite fusion image; A scaling module, used for scaling the satellite fusion image by bicubic interpolation to obtain a satellite thumbnail; An extraction module is used to perform image segmentation on the satellite thumbnail according to an image watershed algorithm to obtain N image blocks, and then extract cloud features of each pixel in the image block, wherein the cloud features include at least one of the following: gray value features, morphological features, color space features, and texture features, and N is a positive integer; An evaluation module, used to evaluate the cloud amount in each image block based on the cloud amount characteristics of each pixel point in the image block, so as to determine a rough judgment result of the cloud amount of the satellite image thumbnail; Wherein, the device is also used for: By constructing an energy function of the satellite thumbnail, an energy state of each pixel in the satellite thumbnail is obtained; Select P seed pixels in the satellite thumbnail, start simulating the diffusion of water from the seed pixels, and flow from the pixel with a higher energy function value to the adjacent pixel with a lower energy value; When different seed pixel points flow in the same direction, the water level of the meeting pixel point will form a watershed pixel point, and each watershed pixel point in the satellite thumbnail image is obtained; Obtaining N image blocks after image segmentation according to the pixel areas divided by each of the watershed pixel points; Wherein, the device is also used for: For each of the image blocks, cloud pixel points that meet a preset cloud feature standard are obtained from each of the pixel points; wherein the preset cloud feature standard includes at least one of the following: a gray value feature standard, a morphological feature standard, a color space feature standard, and a texture feature standard; Traversing the cloud pixel points through each pixel point in the image block in the row direction and the column direction until traversing to a non-cloud pixel point or an image boundary, so as to obtain a cloud coverage area in each image block; Determining a rough cloud amount determination result of the satellite image thumbnail according to the cloud coverage area in each image block; Wherein, the device is also used for: Determining a rough cloud amount determination result for each of the image blocks according to the proportion of the cloud coverage area in each of the image blocks; The cloud amount rough discrimination results of each of the image blocks are combined to obtain the cloud amount rough discrimination result of the satellite image thumbnail; The device is also used for: Acquire an initial satellite image that meets a preset cloud cover assessment standard; wherein the preset cloud cover assessment standard includes: a cloud cover evaluation standard and a cloud cover assessment grade; The initial satellite image is subjected to absolute radiometric correction to obtain a corrected satellite image to be processed.

4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the program, the method for roughly distinguishing cloud cover based on satellite image thumbnails as described in any one of claims 1 to 2 is implemented.

5. A non-transitory computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the method for roughly distinguishing cloud cover based on satellite image thumbnails as described in any one of claims 1 to 2 is implemented.

Images