Marine radar oil spill target segmentation method and system

Abstract

The invention discloses a marine radar oil spill target segmentation method and system, and relates to the technical field of target segmentation. The method is characterized by comprising the following steps: preprocessing a shipborne radar image; the preprocessing comprises the step of carrying out convolution on the shipborne radar image by adopting a Laplacian operator; an Otsu algorithm is adopted to segment the image after convolution, and same-frequency interference noise is extracted; smoothing the same-frequency interference noise by means of mean filtering to obtain a preliminary smoothed image; segmenting the preliminary smooth image by using a gray threshold algorithm and a pixel area threshold method, and extracting block noise; smoothing the block noise by adopting a dynamic window and distance weighted median filtering to obtain a denoised image; and segmenting the denoised image by using an improved Phansalkar local threshold method, and extracting an oil film. According to the method, the oil spill target features are more prominent by effectively suppressing noise, so that the accuracy of oil spill detection is remarkably improved, and the actual requirements of effective region extraction in the field of oil spill detection in marine radar images are met.

Description

Technical Field

[0001] The present invention relates to the technical field of target segmentation, and in particular to a method and system for segmenting marine radar oil spill targets. Background Art

[0002] The fast noise reduction and segmentation method of marine radar oil spill data is mainly used in marine radar systems to identify oil spill areas on the sea surface and process related data. After an oil spill accident occurs, it is crucial to be able to monitor and judge the scope and spread of the oil spill in a timely and accurate manner for environmental protection and accident emergency response.

[0003] Radar systems usually obtain a large amount of data through sea surface monitoring and use it for oil spill detection. However, due to the characteristics of radar data itself and the influence of the external environment, the data often contains a lot of noise, which brings challenges to subsequent data analysis and oil spill detection. Summary of the invention

[0004] In view of the above problems, the present invention proposes a method and system for segmenting oil spill targets using a marine radar, in an effort to solve or alleviate one or more of the above problems.

[0005] According to one aspect of the present invention, a method for segmenting a marine radar oil spill target is proposed, the method comprising:

[0006] Acquire shipborne radar images;

[0007] The shipborne radar image is preprocessed; the preprocessing includes: convolving the shipborne radar image with a Laplace operator; segmenting the convolved image with an Otsu algorithm to extract the same-frequency interference noise; smoothing the same-frequency interference noise with a mean filter to obtain a preliminary smoothed image; segmenting the preliminary smoothed image with a grayscale threshold algorithm and a pixel area threshold method to extract block noise; smoothing the block noise with a dynamic window and a distance-weighted median filter to obtain a denoised image;

[0008] The improved Phansalkar local threshold method is used to segment the denoised image and extract the oil film.

[0009] Furthermore, the preprocessing step of smoothing the block noise by using a dynamic window and a distance-weighted median filter includes:

[0010] For each noise point of block noise, a dynamic local window of size n×n is selected with the noise point as the center, where n is greater than or equal to 3 and is an odd number; if the number of non-noise points in the dynamic local window of size n×n is greater than the number of noise points, noise point suppression is performed; otherwise, the size of the dynamic local window is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points;

[0011] When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight:

[0012]

[0013] Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window;

[0014] The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

[0015] Furthermore, the segmentation threshold in the improved Phansalkar local threshold method is calculated according to the following formula:

[0016]

[0017] In the formula, m represents the local mean; p and q are constants; s represents the local standard deviation; k is a user-defined parameter; Represents the maximum value of all local window variances.

[0018] Furthermore, the method also includes: after preprocessing the shipborne radar image, using a grayscale correction matrix to perform grayscale correction on the preprocessed image; specifically including: calculating the mean of the pixels in the denoised image in the row direction, and all the means form a column vector C1; copying the column vector C1 by column to generate a grayscale correction matrix of the same size as the denoised image; and then subtracting the denoised image from the grayscale correction matrix to obtain a grayscale corrected image.

[0019] Furthermore, the method also includes: after grayscale correction of the image, using a local contrast enhancement method to enhance the features of the oil film in the grayscale corrected image; specifically including: decomposing the image into sub-windows of equal size; calculating the grayscale histogram of each sub-window; adjusting the histogram in each sub-window: evenly distributing other grayscale levels to pixels with a probability greater than a threshold β; equalizing the adjusted histogram; performing bilinear interpolation operations between sub-windows to generate a final contrast enhanced image.

[0020] According to another aspect of the present invention, a marine radar oil spill target segmentation system is also provided, the system comprising:

[0021] An image acquisition module configured to acquire a shipborne radar image;

[0022] A preprocessing module is configured to preprocess the shipborne radar image; the preprocessing includes: convolving the shipborne radar image with a Laplace operator; segmenting the convolved image with an Otsu algorithm to extract the same-frequency interference noise; smoothing the same-frequency interference noise with a mean filter to obtain a preliminary smoothed image; segmenting the preliminary smoothed image with a grayscale threshold algorithm and a pixel area threshold method to extract block noise; and smoothing the block noise with a dynamic window and a distance-weighted median filter to obtain a denoised image;

[0023] The segmentation module is configured to segment the denoised image using an improved Phansalkar local threshold method to extract the oil film.

[0024] Furthermore, the use of dynamic window and distance weighted median filtering to smooth block noise in the preprocessing module includes:

[0025] For each noise point of block noise, a dynamic local window of size n×n is selected with the noise point as the center, where n is greater than or equal to 3 and is an odd number; if the number of non-noise points in the dynamic local window of size n×n is greater than the number of noise points, noise point suppression is performed; otherwise, the size of the dynamic local window is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points;

[0026] When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight:

[0027]

[0028] Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window;

[0029] The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

[0030] Furthermore, the segmentation threshold in the improved Phansalkar local threshold method in the segmentation module is calculated according to the following formula:

[0031]

[0032] In the formula, m represents the local mean; p and q are constants; s represents the local standard deviation; k is a user-defined parameter; Represents the maximum value of all local window variances.

[0033] Furthermore, the system also includes: a grayscale correction module, which is configured to use a grayscale correction matrix to perform grayscale correction on the preprocessed image; specifically including: calculating the mean of the pixels in the denoised image in the row direction, and all the means form a column vector C1; copying the column vector C1 by column to generate a grayscale correction matrix of the same size as the denoised image; and then subtracting the denoised image from the grayscale correction matrix to obtain a grayscale corrected image.

[0034] Furthermore, the system also includes: a feature enhancement module, which is configured to use a local contrast enhancement method to perform feature enhancement on the oil film in the grayscale corrected image; specifically includes: decomposing the image into sub-windows of equal size; calculating the grayscale histogram of each sub-window; performing histogram adjustment in each sub-window: evenly distributing other grayscale levels to pixels with a probability of occurrence greater than a threshold value β; performing equalization processing on the adjusted histogram; performing bilinear interpolation operations between sub-windows to generate a final contrast enhanced image.

[0035] The beneficial technical effects of the present invention are:

[0036] Traditional marine radar oil spill detection methods are easily affected by noise, resulting in unclear distinction between the oil spill area and the background. The present invention proposes a marine radar oil spill target segmentation method and system, which effectively suppresses noise to make the oil spill target features more prominent, thereby significantly improving the accuracy of oil spill detection. The method of the present invention can not only effectively reduce the computational overhead while ensuring high accuracy, but also meet the needs of real-time or near real-time detection, and has significant application advantages. The present invention can effectively suppress noise information in radar images, improve the accuracy of oil spill detection, compress the amount of data, and enhance important features in the image, thereby making offshore oil spill detection more accurate and efficient. The present invention can enhance the robustness of oil spill signals, so that the model or algorithm can still accurately identify the oil spill area when the image is not clear, solving the actual needs of effective area extraction in the field of oil spill detection in marine radar images. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, in which:

[0038] Figure 1 It is a flow chart of a method for segmenting oil spill targets using a marine radar according to an embodiment of the present invention.

[0039] Figure 2 It is a preprocessing flow chart in an embodiment of the present invention.

[0040] Figure 3 is an example diagram of an image after preprocessing in an embodiment of the present invention.

[0041] Figure 4 2 is an example diagram of the grayscale correction process in an embodiment of the present invention.

[0042] Figure 5 is an example graph of grayscale correction generated in an embodiment of the present invention.

[0043] Figure 6 1 is an example diagram of histogram adjustment in the process of image contrast enhancement in an embodiment of the present invention; wherein (a) the portion corresponding to the gray value frequency greater than the threshold β is deleted; and (b) the gray values ​​of the corresponding deleted gray value pixels are evenly distributed.

[0044] Figure 7 This is an example diagram of oil film feature enhancement in an embodiment of the present invention.

[0045] Figure 8 2 is an example diagram of the final segmentation result in an embodiment of the present invention.

[0046] Fig. 9It is a structural schematic diagram of a marine radar oil spill target segmentation system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0048] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, device, apparatus, method or computer program product. Therefore, the present disclosure can be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. In this article, it is to be understood that any number of elements in the drawings is for illustration and not for limitation, and any naming is only for distinction and does not have any limiting meaning.

[0049] The embodiment of the present invention provides a method for segmenting a marine radar oil spill target. Figure 1 As shown, the method includes:

[0050] Step 1: Obtain shipborne radar images;

[0051] Step 2, preprocessing the shipborne radar image; the preprocessing includes: convolving the shipborne radar image with a Laplace operator; segmenting the convolved image with an Otsu algorithm to extract the same-frequency interference noise; smoothing the same-frequency interference noise with a mean filter window to obtain a preliminary smoothed image; segmenting the preliminary smoothed image with a grayscale threshold algorithm and a pixel area threshold method to extract block noise; smoothing the block noise with a dynamic window and a distance-weighted median filter to obtain a denoised image;

[0052] Step 3: Use the grayscale correction matrix to perform grayscale correction on the preprocessed image;

[0053] Step 4: Using a local contrast enhancement method to enhance the features of the oil film in the grayscale corrected image;

[0054] Step 5: Use the improved Phansalkar local threshold method to segment the oil film in the feature enhanced image.

[0055] The method starts with step 1. In step 1, a shipborne radar image is acquired.

[0056] According to an embodiment of the present invention, a plurality of radar images containing oil spills are collected during the same voyage and under the same sea conditions.

[0057] Then, step 2 is performed, in which the shipborne radar image is preprocessed.

[0058] According to an embodiment of the present invention, the preprocessing process is as follows: Figure 2 As shown in the figure, the radar image after coordinate transformation is convolved with the Laplace operator; the convolved image is segmented by the Otsu algorithm to extract the same-frequency interference noise; the same-frequency interference noise is smoothed by the mean filter to obtain a preliminary smoothed image; the preliminary smoothed image is segmented by the gray threshold algorithm and the pixel area threshold method to extract the block noise; the dynamic window and distance weighted median filter are used to smooth the block noise to obtain the denoised image.

[0059] In the prior art, the Otsu algorithm and the pixel area threshold (200) can be used to extract block noise targets; a 21×21 window (the window size is slightly larger than twice the pixel area threshold in step 4) is used to perform median filtering on the block noise targets. The embodiment of the present invention proposes to use a dynamic window and distance-weighted median filtering to suppress each pixel of the block noise, so as to effectively improve the noise smoothing effect. The specific process includes:

[0060] For each noise point of block noise, a dynamic local window of size n×n (n is greater than or equal to 3 and is an odd number) is selected with the noise point as the center; start with a size of 3×3. If the number of non-noise points in the dynamic local window is greater than the number of noise points, noise point suppression can be performed directly; otherwise, the dynamic local window size is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points.

[0061] When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight:

[0062] (1)

[0063] Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window;

[0064] The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

[0065] After the above preprocessing, a preprocessed image of 512×2048 pixels containing the oil spill sample is obtained. Figure 3 shown.

[0066] Then, execute step three, use the grayscale correction matrix to perform grayscale correction on the preprocessed image.

[0067] According to an embodiment of the present invention, a grayscale correction matrix is ​​used to perform global grayscale smoothing on an image to eliminate the influence of background noise. The specific calculation process is as follows: the mean of the denoised image M1 is calculated in the "row" direction to generate a column vector C1; then the mean in C1 is used to fill a grayscale correction matrix M2 of the same specification as M1; then the denoised image M1 is subtracted from the grayscale correction matrix M2 to obtain a grayscale corrected image.

[0068] Examples include Figure 4 The denoised image M1 is a 4×4 matrix, and the mean of each row is 17, 28, 63, and 75, forming a column vector C1. The column vector C1 is expanded into a 4×4 grayscale correction matrix M2, and the value of each row is the same, and the value of each column is C1. The grayscale corrected image example is shown in Figure 5 shown.

[0069] Then, step 4 is performed to enhance the features of the oil film in the grayscale corrected image using a local contrast enhancement method.

[0070] According to an embodiment of the present invention, a CLAHE (Contrast-limited Adaptive Histogram Equalization) algorithm is used to enhance the contrast of an image, thereby enhancing the contrast of areas inside and outside the oil film. Figure 6 The specific implementation process is as follows:

[0071] First, decompose the image into sub-windows of equal size; then, calculate the grayscale histogram of each sub-window; then adjust the histogram in each sub-window: evenly distribute the pixels with a probability greater than the threshold β to other gray levels to avoid over-enhancement of some gray levels; then, equalize the adjusted histogram to enhance the local contrast; finally, perform bilinear interpolation between sub-windows to avoid "block" effects and generate the final contrast-enhanced image. An example of an image after feature enhancement is shown below: Figure 7 shown.

[0072] Then, step five is executed, in which the oil film in the feature-enhanced image is segmented using the improved Phansalkar local threshold method.

[0073] According to an embodiment of the present invention, Niblack proposed a local threshold method for digital image segmentation, as shown below:

[0074] (2)

[0075] Where m is the local mean; is the local standard deviation; k is a user-defined parameter, which takes a negative value. Sauvola modified the method for implementing adaptive document image binarization as follows:

[0076] (3)

[0077] Where R is the dynamic range of the standard deviation and the parameter k obtains positive values. Phansalskar modified the Sauvola method to handle low contrast images as follows:

[0078] (4)

[0079] where p and q are constants. For values ​​of q above a certain value of the local mean, the exponential term becomes negligible. Phansalkar suggests k = 0.25, R = 0.5, p = 2, and q = 10.

[0080] The embodiment of the present invention proposes to improve R to: the maximum value of all local window variances Preferably, k=0.85 and the local window size is 63×63 pixels.

[0081] (5)

[0082] The local window refers to a window of size m×m around the pixel to be calculated, where m×m is, for example, 3×3 or 5×5. The process of calculating the variance of the local window is: first calculate the mean of the grayscale value of the image in the local window, and then calculate the average of the squared difference between the grayscale value and the mean. The example of the result after segmentation is as follows: Figure 8 shown.

[0083] Another embodiment of the present invention provides a marine radar oil spill target segmentation system, such as Fig. 9 As shown, the system includes:

[0084] An image acquisition module 910 configured to acquire a shipborne radar image;

[0085] A preprocessing module 920 is configured to preprocess the shipborne radar image; the preprocessing includes: convolving the shipborne radar image with a Laplace operator; segmenting the convolved image with an Otsu algorithm to extract the same-frequency interference noise; smoothing the same-frequency interference noise with a mean filter to obtain a preliminary smoothed image; segmenting the preliminary smoothed image with a grayscale threshold algorithm and a pixel area threshold method to extract block noise; smoothing the block noise with a dynamic window and a distance-weighted median filter to obtain a denoised image;

[0086] The segmentation module 930 is configured to segment the denoised image using an improved Phansalkar local threshold method to extract the oil film.

[0087] In this embodiment, preferably, the use of dynamic window and distance weighted median filtering to smooth block noise in the preprocessing module 920 includes:

[0088] For each noise point of block noise, a dynamic local window of size n×n is selected with the noise point as the center, where n is greater than or equal to 3 and is an odd number; if the number of non-noise points in the dynamic local window of size n×n is greater than the number of noise points, noise point suppression is performed; otherwise, the size of the dynamic local window is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points;

[0089] When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight:

[0090]

[0091] Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window;

[0092] The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

[0093] In this embodiment, preferably, the segmentation threshold in the improved Phansalkar local threshold method in the segmentation module 930 is calculated according to the following formula:

[0094]

[0095] Where m is the local mean; p and q are constants, and the value of q is higher than a specific value of m; Represents the local standard deviation; k is a user-defined parameter; Represents the maximum value of all local window variances.

[0096] In this embodiment, preferably, the system also includes: a grayscale correction module 940, which is configured to perform grayscale correction on the preprocessed image using a grayscale correction matrix; specifically including: calculating the mean of the pixels in the denoised image in the row direction, and all the means form a column vector C1; copying the column vector C1 by column to generate a grayscale correction matrix of the same size as the denoised image; and then subtracting the denoised image from the grayscale correction matrix to obtain a grayscale corrected image.

[0097] In this embodiment, preferably, the system also includes: a feature enhancement module 950, which is configured to use a local contrast enhancement method to perform feature enhancement on the oil film in the grayscale corrected image; specifically including: decomposing the image into sub-windows of equal size; calculating the grayscale histogram of each sub-window; performing histogram adjustment in each sub-window: evenly distributing other gray levels to pixels with a probability of occurrence greater than a threshold value β; performing equalization processing on the adjusted histogram; performing bilinear interpolation operations between sub-windows to generate a final contrast enhanced image.

[0098] For the undetailed parts of the marine radar oil spill target segmentation system according to the embodiment of the present invention, please refer to the above detailed description of the method embodiment.

[0099] The method of the present invention can be executed in an electronic device. The electronic device can be any device with storage and computing capabilities, which can be implemented as a server, a workstation, etc., or as a personal computer such as a desktop computer or a notebook computer, or as a terminal device such as a mobile phone, a tablet computer, a smart wearable device, an Internet of Things device, but is not limited thereto.

[0100] The electronic device may include: a processor, a memory, an input / output interface, a communication interface and a bus. The processor, the memory, the input / output interface and the communication interface are connected to each other through the bus in the electronic device. The processor may be implemented in the form of a general-purpose CPU, a microprocessor, an application-specific integrated circuit, or one or more integrated circuits, etc., for executing relevant programs to implement the technical solutions provided in the embodiments of this specification. The memory may be implemented in the form of ROM, RAM, static storage devices, dynamic storage devices, etc. The memory may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory and called and executed by the processor. The input / output interface is used to connect the input / output module to realize information input and output. The input / output / module may be configured in the electronic device as a component, or it may be externally connected to the electronic device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc. The communication interface is used to connect the communication module to realize the communication interaction between the electronic device and other devices. The communication module may realize communication by wire or by wireless. A bus comprises a pathway that transfers information between the various components of an electronic device.

[0101] The embodiment of the present invention also provides a non-transitory readable storage medium, which stores instructions, and the instructions are used to enable the electronic device to perform a method according to an embodiment of the present invention. The readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be a computer-readable instruction, a data structure, a module of a program, or other data. Examples of readable storage media include, but are not limited to: phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape disk storage, etc.

[0102] It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present application. As shown in the present specification, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "include one..." do not exclude the presence of other identical elements in the process, method or device including the elements.

[0103] 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 replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A method for segmenting marine radar oil spill targets, characterized in that: include: Acquire shipborne radar images; The shipborne radar image is preprocessed; the preprocessing includes: convolving the shipborne radar image with a Laplace operator; segmenting the convolved image with an Otsu algorithm to extract the same-frequency interference noise; smoothing the same-frequency interference noise with a mean filter to obtain a preliminary smoothed image; segmenting the preliminary smoothed image with a grayscale threshold algorithm and a pixel area threshold method to extract block noise; smoothing the block noise with a dynamic window and a distance-weighted median filter to obtain a denoised image; The improved Phansalkar local threshold method is used to segment the denoised image and extract the oil film.

2. A method for segmenting marine radar oil spill targets according to claim 1, characterized in that: The method of smoothing block noise by using dynamic window and distance weighted median filtering in the preprocessing includes: For each noise point of block noise, a dynamic local window of size n×n is selected with the noise point as the center, where n is greater than or equal to 3 and is an odd number; if the number of non-noise points in the dynamic local window of size n×n is greater than the number of noise points, noise point suppression is performed; otherwise, the size of the dynamic local window is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points; When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight: ， Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window; The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

3. A method for segmenting marine radar oil spill targets according to claim 1, characterized in that: The segmentation threshold in the improved Phansalkar local threshold method is calculated according to the following formula: ， Where m represents the local mean; p and q are constants; Represents the local standard deviation; k is a user-defined parameter; Represents the maximum value of all local window variances.

4. The method for segmenting marine radar oil spill targets according to claim 1, characterized in that: The method also includes: after preprocessing the shipborne radar image, using a grayscale correction matrix to perform grayscale correction on the preprocessed image; specifically including: calculating the mean of pixel points in the denoised image in the row direction, all the means form a column vector C1; copying the column vector C1 by column to generate a grayscale correction matrix of the same size as the denoised image; and then subtracting the denoised image from the grayscale correction matrix to obtain a grayscale corrected image.

5. A method for segmenting marine radar oil spill targets according to claim 4, characterized in that: The method further comprises: after grayscale correction is performed on the image, a local contrast enhancement method is used to enhance the features of the oil film in the grayscale corrected image; specifically, the method comprises: decomposing the image into sub-windows of equal size; calculating the grayscale histogram of each sub-window; adjusting the histogram in each sub-window: evenly distributing other grayscale levels to pixels with an appearance probability greater than a threshold value β; performing equalization processing on the adjusted histogram; and performing bilinear interpolation operations between the sub-windows to generate a final contrast enhanced image.

6. A marine radar oil spill target segmentation system, characterized in that: include: An image acquisition module configured to acquire a shipborne radar image; A preprocessing module, configured to preprocess the shipborne radar image; The preprocessing includes: using the Laplace operator to convolve the shipborne radar image; using the Otsu algorithm to segment the convolved image and extract the same-frequency interference noise; using the mean filter to smooth the same-frequency interference noise to obtain a preliminary smoothed image; using the gray threshold algorithm and the pixel area threshold method to segment the preliminary smoothed image and extract block noise; using the dynamic window and distance weighted median filter to smooth the block noise and obtain a denoised image; The segmentation module is configured to segment the denoised image using an improved Phansalkar local threshold method to extract the oil film.

7. The marine radar oil spill target segmentation system according to claim 6, characterized in that: The method of smoothing block noise by using dynamic window and distance weighted median filtering in the preprocessing module includes: For each noise point of block noise, a dynamic local window of size n×n is selected with the noise point as the center, where n is greater than or equal to 3 and is an odd number; if the number of non-noise points in the dynamic local window of size n×n is greater than the number of noise points, noise point suppression is performed; otherwise, the size of the dynamic local window is set to (n+2)×(n+2), and the above process is repeated until the number of non-noise points in the dynamic local window is no greater than the number of noise points; When the number of non-noise points in the dynamic local window is greater than the number of noise points, the new grayscale value f(i, j) of each non-noise point is calculated according to the distance weight: ， Among them, cell() is the upward rounding function, (i, j) is the position of the non-noise point in the local window, is the original gray value of the non-noise point (i, j) in the local window; The newly calculated non-noise point grayscale values ​​f(i, j) are sorted, and the non-noise point grayscale value f(i, j) corresponding to the sorted median point is replaced by the noise point grayscale value at the center of the local window.

8. The marine radar oil spill target segmentation system according to claim 6, characterized in that: The segmentation threshold in the improved Phansalkar local threshold method in the segmentation module is calculated according to the following formula: ， Where m represents the local mean; p and q are constants; Represents the local standard deviation; k is a user-defined parameter; Represents the maximum value of all local window variances.

9. The marine radar oil spill target segmentation system according to claim 6, characterized in that: The system also includes: a grayscale correction module, which is configured to use a grayscale correction matrix to perform grayscale correction on the preprocessed image; specifically, the module includes: calculating the mean of the pixels in the denoised image in the row direction, and all the means form a column vector C1; copying the column vector C1 by column to generate a grayscale correction matrix of the same size as the denoised image; and then performing a subtraction operation on the denoised image and the grayscale correction matrix to obtain a grayscale corrected image.

10. The marine radar oil spill target segmentation system according to claim 9, characterized in that: The system also includes: a feature enhancement module, which is configured to use a local contrast enhancement method to enhance the features of the oil film in the grayscale corrected image; specifically includes: decomposing the image into sub-windows of equal size; calculating the grayscale histogram of each sub-window; adjusting the histogram in each sub-window: evenly distributing other grayscale levels to pixels with a probability greater than a threshold β; performing equalization processing on the adjusted histogram; performing bilinear interpolation operations between the sub-windows to generate a final contrast enhanced image.

Images