Oceanic internal wave sar image texture enhancement method, device, equipment and medium

Abstract

The present application relates to the technical field of SAR image processing, and particularly relates to a sea internal wave SAR image texture enhancement method, device, equipment and medium, wherein the method comprises: acquiring sea internal wave SAR image real data to be enhanced; performing filtering and denoising on the sea internal wave SAR image real data to obtain a filtered image; extracting an image structure layer based on the filtered image; extracting an image texture layer based on the filtered image and the extracted image structure layer; enhancing the extracted image texture layer; and reconstructing an enhanced sea internal wave SAR image based on the image structure layer and the enhanced image texture layer. The present application can enhance the texture details of the sea internal wave SAR image according to the characteristics of the sea internal wave SAR image.

Description

Technical Field

[0001] This invention relates to the field of SAR image processing technology, and in particular to a method, apparatus, electronic device, and storage medium for enhancing the texture of ocean internal wave SAR images. Background Technology

[0002] Synthetic Aperture Radar (SAR) possesses advantages such as all-weather, all-time capability, long-range coverage, wide coverage, and high resolution, making it a primary technical means for observing and studying internal ocean wave phenomena. However, the imaging quality of SAR for internal ocean wave phenomena is not stable enough, and SAR images of internal ocean waves often lack clear texture details, which is not conducive to observation and analysis.

[0003] Traditional SAR image enhancement methods are often difficult to adapt to ocean internal wave SAR images, resulting in poor image texture enhancement. Summary of the Invention

[0004] Given that existing SAR image enhancement technologies are not specifically tailored to the characteristics of ocean internal wave SAR images, resulting in unsatisfactory image texture enhancement effects, this invention provides a method, apparatus, electronic device, and storage medium for enhancing the texture details of ocean internal wave SAR images, which can enhance the texture details of ocean internal wave SAR images according to their characteristics.

[0005] In a first aspect, embodiments of the present invention provide a method for enhancing the texture of an ocean internal wave SAR image, comprising:

[0006] Acquire real data of the internal wave SAR image of the ocean to be enhanced;

[0007] The real data of the ocean internal wave SAR image is filtered and denoised to obtain the filtered image;

[0008] Based on the filtered image, extract the image structure layer;

[0009] Based on the filtered image and the extracted image structure layer, the image texture layer is extracted;

[0010] Enhance the extracted image texture layer;

[0011] Based on the image structure layer and the enhanced image texture layer, an enhanced ocean internal wave SAR image is reconstructed.

[0012] Optionally, the filtering and noise reduction of the real data of the ocean internal wave SAR image includes:

[0013] The grayscale value of each pixel in the real data of the acquired ocean internal wave SAR image is reconstructed using the nonlocal mean filtering method.

[0014] Optionally, grayscale reconstruction is performed on a pixel in the acquired real data of the ocean internal wave SAR image, including:

[0015] Let the pixel whose grayscale value needs to be reconstructed be the target pixel;

[0016] Based on the coordinates of the target pixel and the preset search window side length, sliding window side length and smoothing parameters, the weights of all pixels in the search window when reconstructing the target pixel are calculated and normalized.

[0017] The reconstructed grayscale value of the target pixel is calculated based on the pixel values ​​of all pixels within the search window and the normalized weights.

[0018] Optionally, the step of calculating the weights of all pixels within the search window in reconstructing the target pixel based on the coordinates of the target pixel and preset search window side length, sliding window side length, and smoothing parameters includes:

[0019] Based on the coordinates of the target pixel and the preset side length of the search window, a square region centered on the target pixel and with the side length of the search window is determined and set as the search window.

[0020] Perform the following operation on each pixel in the search window:

[0021] Set a pixel within the search window as the search pixel, define a square region centered on the search pixel with the side length of the sliding window as the sliding window, and set it as the sliding window. Define a square region centered on the target pixel with the side length of the sliding window as the target window.

[0022] Calculate the weighted Euclidean distance between the target window and the sliding window;

[0023] Based on the calculated weighted Euclidean distance and preset smoothing parameters, the weights of the search pixels are calculated when reconstructing the target pixels.

[0024] Optionally, the step of extracting the image structure layer based on the filtered image includes:

[0025] Let the number of iterations t = 0, and let the initial image structure layer... This represents the filtered image;

[0026] The expression for calculating the four nonlinear terms corresponding to all pixels in the current image structure layer is as follows:

[0027]

[0028]

[0029]

[0030]

[0031] Among them, G σ This represents a Gaussian filter with a standard deviation of σ, where σ is a preset scaling parameter, and * is the convolution operator. and These represent the partial derivatives of image structure layer S in the x and y directions, respectively, with the subscript q indicating the q-th pixel in the corresponding image structure layer S, and u... xq w xq u yq and w yq These are used to represent the four nonlinear terms of the q-th pixel, ε s This is the preset sharpness parameter;

[0032] Based on the four nonlinear terms corresponding to all pixels, calculate the weight matrix L for the current iteration number. t The expression is:

[0033]

[0034] Among them, C x and C y U represents the Topplitz matrix obtained by the forward difference discrete gradient operator in the x and y directions, respectively. x U y W x and W y It is a diagonal matrix, with diagonal elements U. x (i, i) = u xi U y (i, i) = u yi W x (i, i) = w xi and W y (i, i) = w yi , i represents the i-th pixel, and the superscript T indicates taking the transpose of the matrix;

[0035] Based on the filtered image and the weight matrix L at the current iteration number t The image structure layer for the next iteration is calculated using the following expression:

[0036]

[0037] Where 1 represents the identity matrix, Let S be the vector representation of the image structure layer S for the next iteration. The filtered image The vector representation of λ, where λ is a preset intensity parameter;

[0038] Update the loop iteration count t = t + 1. If the loop iteration count t reaches the preset structural layer extraction iteration count N, iter If the condition is met, the extracted image structure layer is obtained; otherwise, the process returns to the step of calculating the four nonlinear terms corresponding to all pixels in the current image structure layer.

[0039] Optionally, the extraction of the image texture layer uses the following expression:

[0040]

[0041] Where (i, j) represents the coordinates of a pixel; T(i, j) represents an element in the extracted image texture layer T. Represents the image based on filtering The elements in S(i,j) represent the extracted elements in the image structure layer S;

[0042] The enhancement of the extracted image texture layer includes:

[0043] Based on the preset texture enhancement parameter α, the extracted image texture layer is enhanced, as expressed by:

[0044] T′(i,j)=[T(i,j)] α

[0045] Where T′(i,j) represents an element in the enhanced image texture layer T′, and the texture enhancement parameter α is greater than 1.

[0046] Optionally, the enhanced ocean internal wave SAR image is reconstructed based on the image structure layer and the enhanced image texture layer, using the following expression:

[0047] I′(i,j)=S(i,j)×T′(i,j)

[0048] Where I′(i,j) represents an element in the enhanced ocean internal wave SAR image I′.

[0049] Secondly, embodiments of the present invention also provide an ocean internal wave SAR image texture enhancement device, comprising:

[0050] The acquisition module is used to acquire real data of the ocean internal wave SAR image to be enhanced;

[0051] The filtering module is used to filter and reduce noise on the real data of the ocean internal wave SAR image to obtain the filtered image;

[0052] The first extraction module is used to extract image structure layers based on the filtered image;

[0053] The second extraction module is used to extract the image texture layer based on the filtered image and the extracted image structure layer;

[0054] An enhancement module is used to enhance the extracted image texture layer;

[0055] The reconstruction module is used to reconstruct an enhanced ocean internal wave SAR image based on the image structure layer and the enhanced image texture layer.

[0056] Thirdly, embodiments of the present invention also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the ocean internal wave SAR image texture enhancement method described in any embodiment of this specification.

[0057] Fourthly, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the ocean internal wave SAR image texture enhancement method described in any embodiment of this specification.

[0058] This invention provides a method, apparatus, electronic device, and storage medium for enhancing the texture of ocean internal wave SAR images. The invention first filters and reduces noise from the actual data of the ocean internal wave SAR image, then extracts the image texture layer based on the filtered image, enhances the image texture layer, and finally obtains a texture-enhanced ocean internal wave SAR image. By performing a filtering operation before texture enhancement, this invention effectively avoids the amplification effect of texture enhancement on image noise, significantly improving image quality. Furthermore, compared with traditional SAR image enhancement methods, it is more targeted to the characteristics of ocean internal wave SAR images, and can effectively enhance the texture features in ocean internal wave SAR images with a generally dark background. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 This is a schematic diagram of the steps of a marine internal wave SAR image texture enhancement method according to an embodiment of the present invention;

[0061] Figure 2 This is a flowchart of another ocean internal wave SAR image texture enhancement method provided by an embodiment of the present invention;

[0062] Figure 3(a) is an internal ocean SAR image that needs to be enhanced;

[0063] Figure 3(b) is a texture-enhanced SAR image of the ocean internal wave from Figure 3(a);

[0064] Figure 4 This is a hardware architecture diagram of an electronic device provided in an embodiment of the present invention;

[0065] Figure 5 This is a structural diagram of a marine internal wave SAR image texture enhancement device provided in an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0067] Traditional SAR image enhancement methods mainly include spatial methods such as histogram equalization and frequency domain methods such as wavelet transform, which have a certain image enhancement effect on general SAR images. However, ocean internal wave SAR images are characterized by high speckle noise intensity and a generally dark background. Spatial methods such as histogram equalization will cause over-enhancement of ocean internal wave SAR images while losing a large amount of texture detail information. Frequency domain methods such as wavelet transform, while enhancing the texture of ocean internal wave SAR images, will destroy other ocean features in the image and increase the intensity of image speckle noise, resulting in a decrease in image quality. Therefore, existing SAR image enhancement techniques are generally difficult to adapt to ocean internal wave SAR images, resulting in poor texture enhancement effects. In view of this, the present invention provides a method, device, electronic device, and storage medium for enhancing the texture details of ocean internal wave SAR images, which can enhance the texture details of ocean internal wave SAR images according to their characteristics.

[0068] The following describes the specific implementation of the above concept.

[0069] Please refer to Figure 1 This invention provides a method for enhancing the texture of ocean internal wave SAR images, the method comprising:

[0070] Step 100: Obtain real data of the ocean internal wave SAR image to be enhanced;

[0071] Step 102: Filter and denoise the acquired real data of the ocean internal wave SAR image to obtain the filtered image;

[0072] Step 104: Based on the filtered image obtained in step 102, extract the image structure layer;

[0073] Step 106: Based on the filtered image obtained in step 102 and the image structure layer extracted in step 104, extract the image texture layer;

[0074] Step 108: Enhance the extracted image texture layer;

[0075] Step 110: Based on the image structure layer extracted in step 104 and the image texture layer enhanced in step 108, the enhanced ocean internal wave SAR image is reconstructed.

[0076] In this embodiment of the invention, the acquired real data of ocean internal wave SAR image is first filtered and denoised to reduce noise interference in the data. Then, the image texture layer is extracted based on the filtered image, and the image texture layer is enhanced to strengthen the texture details in the image. Finally, an enhanced texture ocean internal wave SAR image is obtained. This invention performs a filtering operation before the texture enhancement operation, which can effectively avoid the amplification effect of the texture enhancement operation on image noise and significantly improve image quality. Furthermore, the image texture layer enhancement method adopted in this invention is more targeted to the characteristics of ocean internal wave SAR images compared with traditional SAR image enhancement methods, and can effectively enhance the texture features in ocean internal wave SAR images with a generally dark background.

[0077] Optionally, for step 102, filtering and denoising the real data of the ocean internal wave SAR image includes:

[0078] The grayscale value of each pixel in the real data of the acquired ocean internal wave SAR image is reconstructed using the nonlocal mean filtering method.

[0079] Considering the characteristics of high speckle noise intensity and overall dark background in ocean internal wave SAR images, nonlocal mean filtering helps to suppress image noise while preserving image texture details. In other embodiments, other filtering methods can also be used for noise reduction.

[0080] Further, in step 102, grayscale value reconstruction is performed on any pixel in the acquired ocean internal wave SAR image real data, including:

[0081] Let the pixel whose grayscale value needs to be reconstructed be the target pixel;

[0082] Based on the coordinates of the target pixel and the preset search window side length, sliding window side length and smoothing parameters, the weights of all pixels in the search window when reconstructing the target pixel are calculated and normalized.

[0083] The reconstructed grayscale value of the target pixel is calculated based on the pixel values ​​of all pixels within the search window and the normalized weights.

[0084] Further, the step of calculating the weights of all pixels within the search window in reconstructing the target pixel based on the coordinates of the target pixel and the preset search window side length, sliding window side length, and smoothing parameters includes:

[0085] Based on the coordinates of the target pixel i and the preset search window side length L s Determine the search window with the target pixel i as the center and the side length L. s Set up a square region with sides of length [length missing] as the search window;

[0086] Perform the following operation on each pixel in the search window:

[0087] Set a pixel within the search window as search pixel j, and define a square region centered at search pixel j with side length equal to the side length of the sliding window, as sliding window N. j Define a square region centered on the target pixel with side length equal to the side length of the sliding window, and denote it as the target window N. i ;

[0088] The weighted Euclidean distance between the target window and the sliding window is calculated using the following expression: Where I(N) i ) represents the target window N i The pixel value matrix within, I(N) j ) represents the sliding window N j The pixel value matrix within;

[0089] Based on the calculated weighted Euclidean distance and preset smoothing parameters, the weights of the search pixels when reconstructing the target pixels are calculated, and the expression is:

[0090]

[0091] Where w(i,j) represents the weight of the search pixel j when reconstructing the target pixel i, and d(i,j) represents the weighted Euclidean distance between the target pixel i and the search pixel j.

[0092] Optionally, step 104 further includes:

[0093] Let the number of iterations t = 0, and let the initial image structure layer... This represents the filtered image; the current image structure layer S can be considered as a function related to t.

[0094] The four nonlinear terms corresponding to all pixels in the current image structure layer S are calculated as follows:

[0095]

[0096]

[0097]

[0098]

[0099] Among them, G σ This represents a Gaussian filter with a standard deviation of σ, where σ is a preset scale parameter that controls the maximum scale of subsequently extracted texture layer elements. The larger σ is, the larger the maximum scale of the texture layer elements. σ is generally an integer in the range [1, 5]. * represents the convolution operator. and Let represent the partial derivatives of image structure layer S in the x and y directions, respectively, and let the subscript q represent the q-th pixel in the corresponding image structure layer S, i.e., (). q This represents the q-th result value corresponding to the q-th pixel after performing the operation within the parentheses on the image structure layer S. This represents the q-th result value of the partial derivative of the image structure layer S in the x-direction. Let u represent the q-th result value of the partial derivative of the image structure layer S in the y-direction. xq w xq u yq and w yq These four nonlinear terms, ε, represent the q-th pixel and have no specific physical meaning. s The preset sharpness parameter, ε, controls the sharpness of the structural layers. s The smaller the value, the sharper the structural layer, ε. s The value range of is generally (1e-3, 0.03];

[0100] Based on the four nonlinear terms corresponding to all pixels, calculate the weight matrix L for the current iteration number. t The expression is:

[0101]

[0102] Among them, C x and C y U represents the Toeplitz matrices obtained by the forward difference discrete gradient operator in the x and y directions, respectively. x Uy W x and W y Diagonal matrix, diagonal matrix U x U y W x and W y The diagonal elements in are U x (i, i) = u xi U y (i, i) = u yi W x (i, i) = w xi and W y (i, i) = w yi , i represents the i-th pixel, (i, i) represents the coordinates in the diagonal matrix, and the superscript T indicates taking the transpose of the matrix;

[0103] Based on the filtered image and the weight matrix L at the current iteration number t The image structure layer for the next iteration is calculated using the following expression:

[0104]

[0105] Where 1 represents the identity matrix, Let S be the vector representation of the image structure layer S for the next iteration. The specific method for representing a matrix in vector form is to stack the matrix column by column into column vectors, i.e., The column vector formed by stacking the image structure layers S column by column for the next iteration. The filtered image The vector representation of The filtered image The column vectors are formed by stacking columns one by one. λ is a preset strength parameter that can control the smoothness of the structural layer. The larger λ is, the smoother the structural layer is. The value range of λ is generally (0, 0.05].

[0106] Update the loop iteration count t = t + 1. If the loop iteration count t reaches the preset structural layer extraction iteration count N, iter If the condition is met, the extracted image structure layer is obtained; otherwise, the process returns to calculating the four nonlinear terms corresponding to all pixels in the current image structure layer S, and iterates until t reaches the preset number of extraction iterations N. iter The final image structure layer is then extracted.

[0107] Optionally, step 106 extracts the image texture layer using the following expression:

[0108]

[0109] Where (i, j) represents the coordinates of a pixel; T(i, j) represents an element in the extracted image texture layer T. Represents the image based on filtering The elements in S(i,j) represent the extracted elements in the image structure layer S, all corresponding to pixel points (i,j).

[0110] Further, step 108 includes:

[0111] Based on the preset texture enhancement parameter α, the extracted image texture layer is enhanced, as expressed by:

[0112] T′(i,j)=[T(i,j)]α

[0113] Where T′(i,j) represents the element in the enhanced image texture layer T′, corresponding to the pixel (i,j). The texture enhancement parameter α is greater than 1. If the texture enhancement parameter α is too small, the enhancement effect on the image texture layer may be insignificant. If the texture enhancement parameter α is too large, the image texture layer may be over-enhanced, resulting in image distortion.

[0114] Further, step 110 reconstructs the enhanced ocean internal wave SAR image, using the following expression:

[0115] I′(i,j)=S(i,j)×T′(r,j)

[0116] Where I′(i,j) represents an element in the enhanced ocean internal wave SAR image I′, corresponding to pixel (i,j).

[0117] In some alternative embodiments of the present invention, such as Figure 2 As shown, the present invention provides a method for enhancing the texture of ocean internal wave SAR images, comprising:

[0118] Step 200: Obtain the real data I of the ocean internal wave SAR image to be enhanced and the preset relevant parameters, including: search window side length L. s , Sliding window side length l s Smoothing parameter h, scale parameter σ, intensity parameter λ, sharpness parameter ε s Number of iterations for extracting structural layers N iter and texture enhancement parameter α;

[0119] Step 202: Select a pixel to be reconstructed for grayscale as the target pixel i, based on the coordinates of the target pixel i and the preset search window side length L. s Determine the search window with the target pixel i as the center and the side length L. s Set up a square region with sides of length [length missing] as the search window;

[0120] Step 204: Set a pixel within the search window as the search pixel j, and determine the center of the sliding window with respect to the side length l. s Let N be a square region with side length N, and let it be a sliding window. j Determine the target pixel i as the center and the sliding window side length l. s Let N be a square region with side length N, and set it as the target window. i ;

[0121] Step 206, calculate the target window N i and sliding window N j The weighted Euclidean distance d(i,j) between them is expressed as:

[0122]

[0123] Among them, I(N) i ) represents the target window N i The pixel value matrix within, I(N) j ) represents the sliding window N j The pixel value matrix within;

[0124] Step 208: Based on the calculated weighted Euclidean distance d(i,j) and the preset smoothing parameter h, calculate the weight of the search pixel j when reconstructing the target pixel i, as expressed in the following expression:

[0125]

[0126] Step 210: Determine if all pixels within the search window have been searched. If yes, continue to the next step; otherwise, return to step 204 and change the position of search pixel j until all pixels within the search window have been searched. With the position of search pixel j changed, slide window N. j The location has also changed;

[0127] Step 212: Normalize the weights of all pixels within the search window when reconstructing the target pixel. The expression is:

[0128]

[0129] w nor (i, j) represents the normalized weight of pixel j when reconstructing the target pixel;

[0130] Step 214: Based on the pixel values ​​of all pixels within the search window and the normalized weights, calculate the reconstructed grayscale value of the target pixel i. The expression is:

[0131]

[0132] I(j) represents the pixel value of pixel j;

[0133] Step 216: Determine whether the grayscale value reconstruction of each pixel in the real data of the ocean internal wave SAR image has been completed. If yes, continue to the next step; otherwise, return to step 202 and change the position of the target pixel i. Continue until the reconstructed grayscale values ​​of all pixels in the image have been calculated to obtain the filtered image.

[0134] Step 218, based on the filtered image Extracting image structure layers, including:

[0135] Let the number of iterations t = 0, and let the initial image structure layer... This represents the filtered image;

[0136] The expression for calculating the four nonlinear terms corresponding to all pixels in the current image structure layer is as follows:

[0137]

[0138]

[0139]

[0140]

[0141] Among them, G σ This represents a Gaussian filter with a standard deviation of σ, where σ is a preset scaling parameter, and * is the convolution operator. and These represent the partial derivatives of image structure layer S in the x and y directions, respectively, with the subscript q indicating the q-th pixel in the corresponding image structure layer S, and u... xq w xq u yq and w yq These are used to represent the four nonlinear terms of the q-th pixel, ε s This is the preset sharpness parameter;

[0142] Based on the four nonlinear terms corresponding to all pixels, calculate the weight matrix L for the current iteration number. t The expression is:

[0143]

[0144] Among them, C x and C y U represents the Topplitz matrix obtained by the forward difference discrete gradient operator in the x and y directions, respectively. x U y Wx and W y Diagonal matrix, diagonal matrix U x U y W x and W y The diagonal elements in are U x (i, i) = u xi U y (i, i) = u yi W x (i, i) = w xi and W y (i, i) = w yi , i represents the i-th pixel, (i, i) represents the coordinates in the diagonal matrix, and the superscript T indicates taking the transpose of the matrix;

[0145] Based on the filtered image and the weight matrix L at the current iteration number t The image structure layer for the next iteration is calculated using the following expression:

[0146]

[0147] Where 1 represents the identity matrix, Let S be the vector representation of the image structure layer S for the next iteration. The filtered image The vector representation of λ, where λ is a preset intensity parameter;

[0148] Update the loop iteration count t = t + 1. If the loop iteration count t reaches the preset structural layer extraction iteration count N, iter If the result is positive, the extracted image structure layer is obtained; otherwise, the process of calculating the four nonlinear terms corresponding to all pixels in the current image structure layer is returned.

[0149] Step 220: Based on the filtered image and the extracted image structure layer, extract the image texture layer using the following expression:

[0150]

[0151] Where (i, j) represents the coordinates of a pixel, and T(i, j) represents an element in the extracted image texture layer T. Represents the image based on filtering The elements in S(i,j) represent the extracted elements in the image structure layer S;

[0152] Step 222: Based on the preset texture enhancement parameter α, enhance the extracted image texture layer, as expressed by:

[0153] T′(i,j)=[T(i,j)] α

[0154] Where T′(i,j) represents an element in the enhanced image texture layer T′, and the texture enhancement parameter α is greater than 1;

[0155] Step 224: Based on the image structure layer and the enhanced image texture layer, reconstruct the enhanced ocean internal wave SAR image using the following expression:

[0156] I′(i,j)=S(i,j)×T′(i,j)

[0157] Where I′(i,j) represents an element in the enhanced ocean internal wave SAR image I′.

[0158] In an embodiment of the present invention, the preset relevant parameters involved in a method for enhancing the texture of an ocean internal wave SAR image are shown in Table 1 below:

[0159] Table 1 Relevant Parameters

[0160]

[0161]

[0162] Using the above embodiments, the real data of the ocean internal wave SAR image shown in Figure 3(a) was enhanced, and the result is shown in Figure 3(b). Comparing Figure 3(a) and Figure 3(b), it can be seen that after processing by the method of the present invention, the texture features in the ocean internal wave SAR image are effectively enhanced, the image texture details are clearer, and the image noise is significantly suppressed and not amplified by the texture enhancement operation. The visual effect of the image is significantly improved, and the image quality is greatly improved.

[0163] like Figure 4 , Figure 5 As shown, this embodiment of the invention provides a texture enhancement device for ocean internal wave SAR images. The device embodiment can be implemented through software, hardware, or a combination of both. From a hardware perspective, such as... Figure 4 The diagram shown is a hardware architecture diagram of an electronic device for enhancing the texture of an ocean internal wave SAR image, provided in an embodiment of the present invention. (Except for...) Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing packets. Taking software implementation as an example, such as... Figure 5As shown, a device in a logical sense is formed by the CPU of its electronic device reading the corresponding computer program from non-volatile memory into memory and running it. This embodiment provides a marine internal wave SAR image texture enhancement device, comprising:

[0164] The acquisition module 501 is used to acquire real data of the ocean internal wave SAR image to be enhanced;

[0165] The filtering module 502 is used to filter and reduce noise on the real data of the ocean internal wave SAR image to obtain the filtered image;

[0166] The first extraction module 503 is used to extract image structure layers based on the filtered image;

[0167] The second extraction module 504 is used to extract the image texture layer based on the filtered image and the extracted image structure layer;

[0168] Enhancement module 505 is used to enhance the extracted image texture layer;

[0169] The reconstruction module 506 is used to reconstruct an enhanced ocean internal wave SAR image based on the image structure layer and the enhanced image texture layer.

[0170] In this embodiment of the invention, the acquisition module 501 can be used to execute step 100 in the above method embodiment, the filtering module 502 can be used to execute step 102 in the above method embodiment, the first extraction module 503 can be used to execute step 104 in the above method embodiment, the second extraction module 504 can be used to execute step 106 in the above method embodiment, the enhancement module 505 can be used to execute step 108 in the above method embodiment, and the reconstruction module 506 can be used to execute step 110 in the above method embodiment.

[0171] In a feasible embodiment of the present invention, the filtering and noise reduction of the real data of the ocean internal wave SAR image includes:

[0172] The grayscale value of each pixel in the real data of the acquired ocean internal wave SAR image is reconstructed using the nonlocal mean filtering method.

[0173] In a feasible embodiment of the present invention, grayscale value reconstruction is performed on a pixel in the acquired real data of an ocean internal wave SAR image, including:

[0174] Let the pixel whose grayscale value needs to be reconstructed be the target pixel;

[0175] Based on the coordinates of the target pixel and the preset search window side length, sliding window side length and smoothing parameters, the weights of all pixels in the search window when reconstructing the target pixel are calculated and normalized.

[0176] The reconstructed grayscale value of the target pixel is calculated based on the pixel values ​​of all pixels within the search window and the normalized weights.

[0177] In a feasible embodiment of the present invention, the step of calculating the weights of all pixels within the search window in reconstructing the target pixel based on the coordinates of the target pixel and preset search window side length, sliding window side length, and smoothing parameters includes:

[0178] Based on the coordinates of the target pixel and the preset side length of the search window, a square region centered on the target pixel and with the side length of the search window is determined and set as the search window.

[0179] Perform the following operation on each pixel in the search window:

[0180] Set a pixel within the search window as the search pixel, define a square region centered on the search pixel with the side length of the sliding window as the sliding window, and set it as the sliding window. Define a square region centered on the target pixel with the side length of the sliding window as the target window.

[0181] Calculate the weighted Euclidean distance between the target window and the sliding window;

[0182] Based on the calculated weighted Euclidean distance and preset smoothing parameters, the weights of the search pixels are calculated when reconstructing the target pixels.

[0183] In a feasible embodiment of the present invention, the step of extracting the image structure layer based on the filtered image includes:

[0184] Let the number of iterations t = 0, and let the initial image structure layer... This represents the filtered image;

[0185] The expression for calculating the four nonlinear terms corresponding to all pixels in the current image structure layer is as follows:

[0186]

[0187]

[0188]

[0189]

[0190] Among them, G σThis represents a Gaussian filter with a standard deviation of σ, where σ is a preset scaling parameter, and * is the convolution operator. and These represent the partial derivatives of image structure layer S in the x and y directions, respectively, with the subscript q indicating the q-th pixel in the corresponding image structure layer S, and u... xq w xq u yq and w yq These are used to represent the four nonlinear terms of the q-th pixel, ε s This is the preset sharpness parameter;

[0191] Based on the four nonlinear terms corresponding to all pixels, calculate the weight matrix L for the current iteration number. t The expression is:

[0192]

[0193] Among them, C x and C y U represents the Topplitz matrix obtained by the forward difference discrete gradient operator in the x and y directions, respectively. x U y W x and W y It is a diagonal matrix, with diagonal elements U. x (i, i) = u xi U y (i, i) = u yi W x (i, i) = w xi and W y (i, i) = w yi , i represents the i-th pixel, and the superscript T indicates taking the transpose of the matrix;

[0194] Based on the filtered image and the weight matrix L at the current iteration number t The image structure layer for the next iteration is calculated using the following expression:

[0195]

[0196] Where 1 represents the identity matrix, Let S be the vector representation of the image structure layer S for the next iteration. The filtered image The vector representation of λ, where λ is a preset intensity parameter;

[0197] Update the loop iteration count t = t + 1. If the loop iteration count t reaches the preset structural layer extraction iteration count N, iterIf the condition is met, the extracted image structure layer is obtained; otherwise, the process returns to the step of calculating the four nonlinear terms corresponding to all pixels in the current image structure layer.

[0198] In a feasible embodiment of the present invention, the extraction of the image texture layer is expressed as follows:

[0199]

[0200] Where (i, j) represents the coordinates of a pixel, and T(i, j) represents an element in the extracted image texture layer T. Represents the image based on filtering The elements in S(i,j) represent the extracted elements in the image structure layer S;

[0201] The enhancement of the extracted image texture layer includes:

[0202] Based on the preset texture enhancement parameter α, the extracted image texture layer is enhanced, as expressed by:

[0203] T′(i,j)=[T(i,j)] α

[0204] Where T′(i,j) represents an element in the enhanced image texture layer T′, and the texture enhancement parameter α is greater than 1.

[0205] In a feasible embodiment of the present invention, the enhanced ocean internal wave SAR image is reconstructed based on the image structure layer and the enhanced image texture layer, using the following expression:

[0206] I′(i,j)=S(i,j)×T′(i,j)

[0207] Where I′(i,j) represents an element in the enhanced ocean internal wave SAR image I′.

[0208] It is understood that the structures illustrated in the embodiments of the present invention do not constitute a specific limitation on an ocean internal wave SAR image texture enhancement device. In other embodiments of the present invention, an ocean internal wave SAR image texture enhancement device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0209] The information interaction and execution process between the modules in the above-mentioned device are based on the same concept as the method embodiment of the present invention, and the specific details can be found in the description of the method embodiment of the present invention, and will not be repeated here.

[0210] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements a marine internal wave SAR image texture enhancement method according to any embodiment of this invention.

[0211] This invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform a marine internal wave SAR image texture enhancement method according to any embodiment of this invention.

[0212] Specifically, a system or apparatus equipped with a storage medium may be provided, on which software program code implementing the functions of any of the embodiments described above is stored, and the computer (or CPU or MPU) of the system or apparatus may read and execute the program code stored in the storage medium.

[0213] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0214] Examples of storage media used to provide program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0215] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0216] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion module connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion module execute some and all of the actual operations, thereby realizing the function of any of the above embodiments.

[0217] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0218] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0219] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for enhancing the texture of ocean internal wave SAR images, characterized in that, include: Acquire real data of the internal wave SAR image of the ocean to be enhanced; The real data of the ocean internal wave SAR image is filtered and denoised to obtain the filtered image; Based on the filtered image, extract the image structure layer; Based on the filtered image and the extracted image structure layer, the image texture layer is extracted; Enhance the extracted image texture layer; Based on the image structure layer and the enhanced image texture layer, an enhanced ocean internal wave SAR image is reconstructed. The extracted image texture layer is expressed using the following expression: in,( i , j () represents the coordinates of a pixel; Represents the extracted image texture layer The elements in Represents the image based on filtering The elements in The extracted image structure layer The elements in; the enhancement of the extracted image texture layer includes: Based on preset texture enhancement parameters The extracted image texture layer is enhanced using the following expression: in, This represents the enhanced image texture layer. Elements in the text, texture enhancement parameters Greater than 1; The enhanced ocean internal wave SAR image is reconstructed based on the image structure layer and the enhanced image texture layer, using the following expression: in, This represents the enhanced internal ocean SAR image. The elements in.

2. The ocean internal wave SAR image texture enhancement method according to claim 1, characterized in that, The filtering and noise reduction of the real data of the ocean internal wave SAR image includes: The grayscale value of each pixel in the real data of the acquired ocean internal wave SAR image is reconstructed using the nonlocal mean filtering method.

3. The ocean internal wave SAR image texture enhancement method according to claim 2, characterized in that, Grayscale reconstruction of a single pixel in the acquired ocean internal wave SAR image data includes: Let the pixel whose grayscale value needs to be reconstructed be the target pixel; Based on the coordinates of the target pixel and the preset search window side length, sliding window side length and smoothing parameters, the weights of all pixels in the search window when reconstructing the target pixel are calculated and normalized. The reconstructed grayscale value of the target pixel is calculated based on the pixel values ​​of all pixels within the search window and the normalized weights.

4. The ocean internal wave SAR image texture enhancement method according to claim 3, characterized in that, The calculation of the weights of all pixels within the search window in reconstructing the target pixel, based on the coordinates of the target pixel and the preset search window side length, sliding window side length, and smoothing parameters, includes: Based on the coordinates of the target pixel and the preset side length of the search window, a square region centered on the target pixel and with the side length of the search window is determined and set as the search window. Perform the following operation on each pixel in the search window: Set a pixel within the search window as the search pixel, define a square region centered on the search pixel with the side length of the sliding window as the sliding window, and set it as the sliding window. Define a square region centered on the target pixel with the side length of the sliding window as the target window. Calculate the weighted Euclidean distance between the target window and the sliding window; Based on the calculated weighted Euclidean distance and preset smoothing parameters, the weights of the search pixels are calculated when reconstructing the target pixels.

5. The ocean internal wave SAR image texture enhancement method according to claim 1, characterized in that, The process of extracting the image structure layer based on the filtered image includes: Let the number of iterations t=0, and let the initial image structure layer... ; This represents the filtered image; The expression for calculating the four nonlinear terms corresponding to all pixels in the current image structure layer is as follows: in, The standard deviation is expressed as Gaussian filter, These are preset scale parameters. For convolution operators, and Representing image structure layers respectively In the partial derivatives in the x and y directions, the subscript q represents the q-th pixel in the corresponding image structure layer S. , , and These are used to represent the four nonlinear terms of the q-th pixel. This is the preset sharpness parameter; Calculate the weight matrix for the current iteration number based on the four nonlinear terms corresponding to all pixels. The expression is: in, and Let represent the Toplitz matrices obtained by the forward difference discrete gradient operator in the x and y directions, respectively. , , and It is a diagonal matrix, with diagonal elements as follows: , , and , i represents the i-th pixel, superscript T This indicates taking the transpose of the matrix; Based on the filtered image and the weight matrix at the current iteration number The image structure layer for the next iteration is calculated using the following expression: in, Represents the identity matrix. For the next iteration of the image structure layer The vector representation of The filtered image The vector representation of These are preset strength parameters; Update the loop iteration count t = t + 1. If the loop iteration count t reaches the preset structural layer extraction iteration count... If the condition is met, the extracted image structure layer is obtained; otherwise, the process returns to the step of calculating the four nonlinear terms corresponding to all pixels in the current image structure layer.

6. A texture enhancement device for ocean internal wave SAR images, characterized in that, The apparatus for implementing the method as described in any one of claims 1-5 comprises: The acquisition module is used to acquire real data of the ocean internal wave SAR image to be enhanced; The filtering module is used to filter and reduce noise on the real data of the ocean internal wave SAR image to obtain the filtered image; The first extraction module is used to extract image structure layers based on the filtered image; The second extraction module is used to extract the image texture layer based on the filtered image and the extracted image structure layer; An enhancement module is used to enhance the extracted image texture layer; The reconstruction module is used to reconstruct an enhanced ocean internal wave SAR image based on the image structure layer and the enhanced image texture layer.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the ocean internal wave SAR image texture enhancement method as described in any one of claims 1-5.

8. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed in the computer, it causes the computer to perform the ocean internal wave SAR image texture enhancement method according to any one of claims 1-5.

Images