Sea island waterline detection method based on phase consistency random walk

Abstract

The invention relates to the technical field of coast waterline detection, in particular to an island waterline detection method based on phase consistency random walk, which comprises the following steps of: converting a single-polarized SAR (Synthetic Aperture Radar) image into a phase consistency graph by using a two-dimensional logarithm Gabor filter; an FCM method is adopted to obtain a sea-land membership function, and sea-land global priori is created; constructing a cross-sea bridge marking field based on LSD, and creating a marking field prior of a non-permeable structure connected with an island; introducing a superpixel layer to construct a waterline detection model; and inputting the single-polarized SAR image into the waterline detection model, obtaining each pixel output label, and realizing island waterline detection. The method systematically solves the problems of coherent noise interference in the single-polarized SAR image, difficulty in seed point initialization in a traditional random walk method, insufficient waterline extraction precision caused by a complex island contour, and gray similarity failure in edge detection caused by large ground feature backscattering difference, and can effectively extract the island shoreline of the single-polarized SAR image.

Description

Technical Field

[0001] This invention relates to the field of waterline detection technology, and in particular to a method for detecting the waterline of an island based on phase-consistent random walk. Background Technology

[0002] Currently, island coastline extraction methods for single-polarization SAR images can be classified into seven categories: edge detection (ED) based methods, threshold-based methods, region merging (RM) based methods, partial differential equation (PDE) based methods, deep learning (DL) based methods, Markov random field (MRF) based methods, and superpixel (SP) based methods.

[0003] Edge detection-based methods have the advantage of being simple in principle, easy to understand, and suitable for easy-to-operate waterline extraction, showing good performance in processing SAR images. However, the threshold of edge detection-based methods is difficult to control, and false edges are easily generated.

[0004] Threshold-based methods are advantageous because they are computationally simple and fast, but they are not suitable for detecting complex water edges with uneven contrast.

[0005] The advantage of region merging-based methods lies in their simplicity and ease of implementation. However, automatic selection of seed points is difficult, and they cannot solve the problem of extracting low-contrast regions.

[0006] The advantage of methods based on partial differential equations is that the theory is relatively complete, the evolution is somewhat automatic, and the edge fitting is good. However, it has drawbacks such as sensitivity to the initial contour, slow calculation speed for large images, and the need for manual parameter adjustment.

[0007] The advantages of deep learning-based methods lie in their powerful feature learning capabilities and end-to-end learning process, but they currently have shortcomings in terms of sample size, automation, and generalization ability.

[0008] Markov random field-based methods can make full use of context and structural information, but they are not good at detecting low-contrast and non-uniform regions, and their computation speed is slow.

[0009] Superpixel-based methods use superpixels as primitives, which can greatly reduce computational complexity in subsequent processing, but the accuracy is unstable at complex coastal boundaries.

[0010] Furthermore, in VV single-polarization SAR images, the backscattering coefficients of features such as beaches and forests on islands exhibit spatial non-uniformity, resulting in extremely low contrast in the waterline region. Additionally, the presence of speckle on the complex sea surface causes false detections in existing algorithms. Moreover, the non-uniformity and even partial breaks in island-to-land and island-to-island bridges further complicate research on waterline detection of islands based on SAR images. Therefore, a method is needed to address the interference of coherent noise, difficulties in seed point initialization, insufficient accuracy in waterline extraction, and similarity failure issues in single-polarization SAR images. Summary of the Invention

[0011] To address the aforementioned technical problems, this invention provides a method for detecting the waterline of islands based on phase-consistent random walks. Starting from the phase-consistency domain, in the prior construction, for complex multi-island environments, fuzzy clustering mean (FCM) is used to calculate land-sea membership to provide a global prior. Simultaneously, a method for detecting permeable structures such as cross-sea bridges is proposed based on phase-consistency line segment detection (LSD). A similarity metric is established through phase consistency, and a superpixel layer is added to better utilize image information within the non-local neighborhood. Two priors are introduced to construct the state transition matrix, completing the construction and solution of the RW model based on SAR images.

[0012] The technical means employed in this invention are as follows: A method for detecting the waterline of an island based on phase-consistent random walk includes: converting a single-polarization SAR image into a phase-consistent map using a two-dimensional logarithmic Gabor filter; obtaining the land-sea membership function using the FCM method to create a global prior for land-sea relations; constructing a marker field for cross-sea bridges based on LSD to create a prior for marker fields of impermeable structures connected to the island; introducing a superpixel layer to construct a waterline detection model; and inputting the single-polarization SAR image into the waterline detection model to obtain the output label for each pixel, thereby realizing the detection of the waterline of the island.

[0013] Furthermore, the logarithmic Gabor filter is defined as follows:

[0014] in, and They represent the positions respectively. Image pixels The corresponding angular frequency and direction in polar coordinates and These represent the scale and orientation of the Gabor filter, respectively. and These represent the center frequency and center direction, respectively. Maintain a constant center frequency for different centers. The standard deviation of the Gaussian function in the angular direction; The logarithmic Gabor filter is represented in the spatial domain by the inverse Fourier transform:

[0015] in, and These represent the current logarithmic Gabor wavelet at scale and direction, respectively. and The real and imaginary parts; the image The response after the spatial filter is expressed as:

[0016] in, Represents convolution; Indicates the image at position pixel values, express In terms of scale and direction respectively and The actual response, express In terms of scale and direction respectively and The imaginary part of the response. Furthermore, obtaining the phase consistency map specifically includes: based on the image The phase coherence in each direction, as defined by the response of the spatial filter, is as follows:

[0017] in, , These represent the direction and scale of the log-Gabor filter, respectively. Indicates the interval parameter. Indicates the quantity of scale; For the estimated noise level, It is a constant. This indicates that the included quantity is equal to itself when its value is positive, and zero otherwise; magnitude response. Represented as:

[0018] This indicates that the log-Gabor filter is at a scale of , direction is Phase deviation function at time, Defined as:

[0019]

[0020] in, and They are represented as follows:

[0021]

[0022] Weight function Defined as:

[0023] in, The gain factor is used to control the sharpness of the filter. This is the truncation value. The frequency response extension is expressed as:

[0024] Where N is the total number of scales considered. It is the amplitude at the nth scale at point x. It is the amplitude at which the maximum response is achieved at x.

[0025] Furthermore, the phase consistency map is transformed from image pixels into a phase consistency metric. :

[0026] in, A mathematical function representing phase consistency.

[0027] Furthermore, creating the aforementioned global prior for land and sea specifically includes: using a phase consistency metric as input to obtain the membership degree between land and sea through an FCM classifier, where the membership degree is... Between, definition For phase consistency The membership degree of a node is defined to reflect the two priors of land and sea. This was then introduced as a priori into the waterline detection model.

[0028] Furthermore, the creation of a priori marker field for the impermeable structures connected to the island specifically includes: assuming the existence of a line segment detection function. ,get line segment , To locate the bridge, the following conditions must be met: any two line segments must be in opposite directions and have similar absolute values; the distance between the two line segments must be similar; and each line segment must be long enough to mark the bridge. Defined as:

[0029] in, Represents line segment direction, This represents the calculation of the perpendicular distance between two line segments. This represents the calculation of the length of a line segment. , and These represent the thresholds for angle, distance, and length, respectively.

[0030] Furthermore, the waterline detection model specifically includes: an assumed phase consistency map. have There are 10 nodes, each representing 10 nodes. One element, while based on the phase consistency graph Generated by superpixels Each node is represented by the mean of each superpixel, and the nodes are integrated. Element nodes and superpixel nodes, totaling A weighted graph is generated using nodes, represented as follows: ,in It is a set of nodes. It is a set of edges. The weight matrix has edge connections; in the node set any node Indicates the first indivual Element nodes or superpixel nodes, in the edge set any edge in Indicates at node In the four-neighbor or eight-neighbor domains of the node The connection, It consists of labeled and unlabeled nodes. This represents a set of labeled nodes. , This indicates that there are K labels, and two category labels are needed for land-sea segmentation. Represents a set of uncategorized labels; Assumption The empty domain of a domain element is ,Depend on The set of spatial domains composed of superpixels of a domain is :

[0031] in, Indicates the first The spatial set of the n superpixels, the nth Each superpixel The spatial range is represented as .

[0032] Furthermore, based on the aforementioned phase consistency diagram Connections between nodes in the domain, connections between nodes in the superpixel domain, and The connection relationships between nodes in the domain and corresponding nodes in the superpixel domain are defined, and the weight matrix is ​​defined. for:

[0033] Among them, the weight , , , They are defined as follows:

[0034]

[0035]

[0036]

[0037] in, Phase coherence diagram Nodes in the domain i The value, Phase coherence diagram Nodes in the domain j The value, Indicates the first i The mean value of each superpixel. Indicates the first j The mean of each superpixel; based on Domain elements and superpixels have the same spatial location, and nodes degree Defined as:

[0038] Introducing cluster membership , will node Prior weights Defined as:

[0039] in, These are rule parameters, defining the transition probability as:

[0040] in, and The weights of the two nodes; Based on transition probability From node Walk to the label The probability of node m is:

[0041] Among them, when hour, ,otherwise ; By setting vectors probability The vector expression is:

[0042] in, transition probability matrix Defined as:

[0043] Let be a vector, where:

[0044] For any node Calculate the average probability of reaching Its vector representation is:

[0045] The labels are obtained through calculation: .

[0046] Compared with the prior art, the present invention has the following advantages: This invention provides a method for detecting the waterline of an island based on phase-consistent random walks. It utilizes a two-dimensional logarithmic Gabor filter to transform a single-polarization SAR image into a phase-consistent map; employs the FCM method to obtain the land-sea membership function and create a global prior; constructs a marker field for cross-sea bridges based on LSD to create prior marker fields for impermeable structures connected to the island; and introduces a superpixel layer to construct a waterline detection model. This invention can effectively extract the island coastline from single-polarization SAR images; it systematically solves the problems of coherent noise interference in single-polarization SAR images, difficulty in initializing seed points in traditional random walk methods, insufficient waterline extraction accuracy due to complex island contours, and gray-level similarity failure in edge detection due to large differences in backscattering of ground features.

[0047] The island waterline detection method based on phase consistency random walk provided by this invention makes full use of the constructed phase consistency map, establishes a similarity metric through phase consistency, introduces two priors to construct a state transition matrix, and adopts a "transform domain-prior-walk" model architecture in the phase consistency domain, thereby avoiding the extraction of the two sides of the cross-sea bridge as waterline, and improving the accuracy and practicality of waterline detection. The island waterline detection method based on phase-consistent random walk provided by this invention is not only applicable to island waterline detection, but can also be theoretically extended to multi-target extraction of single-polarization SAR images in other complex scenarios, thus having a certain degree of versatility.

[0048] Based on the above reasons, this invention can be widely applied in fields such as waterline detection. Attached Figure Description

[0049] 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.

[0050] Figure 1 This is a diagram illustrating the architecture of the island waterline detection method based on phase-consistent random walk, as described in this invention.

[0051] Figure 2 This is a comparison image of the extraction of the waterline of an island in an embodiment of the present invention. Detailed Implementation

[0052] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0053] 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 only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0055] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0056] like Figure 1 As shown, this invention provides a method for detecting the waterline of an island based on phase-consistent random walks. In VV single-polarization SAR images, the backscattering coefficients of features such as beaches and forests on islands exhibit spatial non-uniformity, resulting in extremely low contrast in the waterline region. Furthermore, the presence of speckle on the complex sea surface causes false detections in existing algorithms. Additionally, the non-uniformity of bridges between islands and landmasses, and even partial breaks, further complicates SAR image-based island waterline detection research, necessitating an effective solution. Research has shown that obtaining a map from a SAR image through phase consistency can significantly reduce the interference caused by these problems.

[0057] A two-dimensional logarithmic Gabor filter is used to convert a single-polarization SAR image into a phase consistency map. Specifically, as a preferred embodiment of this invention, the logarithmic Gabor filter is defined as follows:

[0058] in, and They represent the positions respectively. Image pixels The corresponding angular frequency and direction in polar coordinates and These represent the scale and orientation of the Gabor filter, respectively. and These represent the center frequency and center direction, respectively. Maintain a constant center frequency for different centers. The standard deviation of the Gaussian function in the angular direction; As a frequency domain filter, the logarithmic Gabor filter is represented in the spatial domain by the inverse Fourier transform:

[0059] in, and These represent the current logarithmic Gabor wavelet at scale and direction, respectively. and The real and imaginary parts; the image The response after the spatial filter is expressed as:

[0060] in, Represents convolution; Indicates the image at position pixel values, express In terms of scale and direction respectively and The actual response, express In terms of scale and direction respectively and The imaginary part of the response. In a specific implementation, as a preferred embodiment of the present invention, obtaining the phase consistency map specifically includes: Based on image The phase coherence in each direction, as defined by the response of the spatial filter, is as follows:

[0061] in, , These represent the direction and scale of the log-Gabor filter, respectively. Indicates the interval parameter. Indicates the number of scales; It is set to 6, meaning the interval is... , It was set to 4. The estimated noise level is typically set to 3. It is a very small constant to avoid being divisible by 0. This indicates that the included quantity is equal to itself when its value is positive, and zero otherwise; magnitude response. Represented as:

[0062] This indicates that the log-Gabor filter is at a scale of , direction is Phase deviation function at time, Defined as:

[0063]

[0064] in, and They are represented as follows:

[0065]

[0066] Weight function Defined as:

[0067] in, The gain factor, which controls the sharpness of the filter, is set to 10. The cutoff value is set to 0.4. The frequency response extension is expressed as:

[0068] Where N is the total number of scales considered. It is the amplitude at the nth scale at point x. It is the amplitude at which the maximum response is located at x.

[0069] Based on the above analysis, the image Phase consistency can be obtained in multiple directions. This means that in position... Existing vector Current literature generally uses multi-directional averaging to describe the phase consistency at a given location, which often introduces blurring at weak edges; the idea is similar to Gaussian smoothing. To highlight phase consistency information in a specific direction, in a preferred embodiment of this invention, the phase consistency map is transformed from image pixels into a phase consistency metric. :

[0070] in, A mathematical function representing phase consistency.

[0071] FCM (Fixed Component Clustering) is a soft clustering method. Its core idea is to determine the membership degree between sample points and cluster centers based on iterative optimization of the objective function. The membership degree is used to represent the relationship between data points, thereby determining the cluster to which each sample point belongs. In this invention, FCM is used to obtain the sea-land membership function, creating a global sea-land prior. Specifically, in a preferred embodiment, a phase consistency metric is used as input to obtain the sea-land membership degree through the FCM classifier. This membership degree... Between, definition For phase consistency The membership degree of a node is defined to reflect the two priors of land and sea. This was then introduced as a priori into the waterline detection model.

[0072] In multi-island scenarios, permeable structures are mainly cross-sea bridges, which can usually be represented by line segments. Therefore, determining their marker field directly involves detecting the corresponding line segments. This section proposes a method for constructing a cross-sea bridge marker field based on LSD (Low Segment Detection). The method involves constructing an LSD-based cross-sea bridge marker field and creating prior marker fields for non-permeable structures connected to islands. Specifically, as a preferred embodiment of this invention, it is assumed that a line segment detection function exists. ,get line segment , To locate the bridge, the following conditions must be met: any two line segments must be in opposite directions and have similar absolute values; the distance between the two line segments must be similar; and each line segment must be long enough to mark the bridge. Defined as:

[0073] in, Represents line segment direction, This represents the calculation of the perpendicular distance between two line segments. This represents the calculation of the length of a line segment. , and These represent the thresholds for angle, distance, and length, respectively.

[0074] In traditional RW (Rapid Wave Wiring), the walking path often connects local pixels. Pixel-based approaches can only compare adjacent pixels pairwise, and in SAR images with speckle noise, they often stop walking when encountering local noise points. Introducing a superpixel layer addresses this problem. Compared to local and global prior connections, superpixel connections better utilize image information within non-local neighborhoods. A waterline detection model is constructed using a superpixel layer; in a preferred embodiment of this invention, a phase consistency map is assumed... have There are 10 nodes, each representing 10 nodes. One element, while based on the phase consistency graph Generated by superpixels Each node is represented by the mean of each superpixel, and the nodes are integrated. Element nodes and superpixel nodes, totaling A weighted graph is generated using nodes, represented as follows: ,in It is a set of nodes. It is a set of edges. The weight matrix has edge connections; in the node set any node Indicates the first indivual Element nodes or superpixel nodes, in the edge set any edge in Indicates at node In the four-neighbor or eight-neighbor domains of the node The connections are categorized into three types: connections between element nodes, connections between superpixel nodes, and connections between superpixels and element nodes. Element node connections are the same as pixel connections in general RW theory. Superpixel node connections simply involve each superpixel center connecting to another superpixel center. The connection between a superpixel and an element node is more special; its superpixel center only connects to... This superpixel region element is connected.

[0075] It consists of labeled and unlabeled nodes. This represents a set of labeled nodes. , This indicates that there are K labels, and two category labels are needed for land-sea segmentation. Represents a set of uncategorized labels; Assumption The empty domain of a domain element is ,Depend on The set of spatial domains composed of superpixels of a domain is :

[0076] in, Indicates the first The spatial set of the n superpixels, the nth Each superpixel The spatial range is represented as .

[0077] Based on the phase consistency diagram Connections between nodes in the domain, connections between nodes in the superpixel domain, and The connection relationships between nodes in the domain and corresponding nodes in the superpixel domain are defined, and the weight matrix is ​​defined. for:

[0078] Among them, the weight , , , They are defined as follows:

[0079]

[0080]

[0081]

[0082] in, Phase coherence diagram Nodes in the domain i The value, Phase coherence diagram Nodes in the domain j The value, Indicates the first i The mean value of each superpixel. Indicates the first j The mean of each superpixel; four weights describe the nodes. and The correlation describes Spatial relationship and similarity measurement between domain elements and superpixels.

[0083] based on Domain elements and superpixels have the same spatial location, and nodes degree Defined as:

[0084] However, relying solely on Completely performing Round-Wave (RW) suffers from a common problem: when a wandering node is too far from the seed point, its wander probability is too low, leading to segmentation errors. Therefore, clustering membership is introduced. , will node Prior weights Defined as:

[0085] in, The rule parameter is a constant greater than 0 and less than 1. The transition probability is defined as:

[0086] in, and The weights of the two nodes; Based on transition probability From node Walk to the label The probability of node m is:

[0087] Among them, when hour, ,otherwise ; By setting vectors probability The vector expression is:

[0088] in, transition probability matrix Defined as:

[0089] Let be a vector, where:

[0090] For any node Calculate the average probability of reaching Its vector representation is:

[0091] The labels are obtained through calculation: .

[0092] By inputting a single-polarization SAR image into a waterline detection model and obtaining the output label for each pixel, waterline detection of islands can be achieved. VV polarization image ,Label , , , , , ; (1) Based on the measure of phase consistency Generate a PCI map and a superpixel map; (2) Generate label vectors , According to the bridge marking field Generate BF; (3) Generate the weight matrix According to the FCM algorithm, ; (4) Calculate the transition matrix respectively and ; (5) Solve the system of linear equations ; (6) Normalized arrival probability ; (7) Obtain the label for each pixel. Output: Output the label for the island waterline for each pixel. .

[0093] Example like Figure 2 As shown, in this embodiment, the waterline detection results of the Ground Truth algorithm, KM-ACM algorithm, HED-Unet algorithm, and SPEC algorithm are compared with the detection results of the algorithm of this invention. Figure 2 In the diagram, (a) is the original image, (b) is the Ground Truth image, (c) is the result of the KM-ACM algorithm, (d) is the result of the HED-Unet algorithm, (e) is the result of the SPEC algorithm, and (f) is the result of the algorithm of this invention. Figure 2 As can be seen, the waterline detected by HED-Unet extends into the land in many places. The extraction results of SPEC differ somewhat from the actual waterline. The extraction results of KM-ACM have a good overall fit, but the recognition effect is slightly worse in some small sharp points. The extraction results of PCRW are basically consistent with the actual waterline and have the best extraction effect.

[0094] 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 or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting the waterline of an island based on phase-consistent random walk, characterized in that, include: A two-dimensional logarithmic Gabor filter is used to convert a single-polarization SAR image into a phase consistency map. The sea-land membership function is obtained using the FCM method, and a global sea-land prior is created. Construct a marker field for the cross-sea bridge based on LSD, and create a priori marker field for the impermeable structures connected to the island; A superpixel layer is introduced to construct a waterline detection model; By inputting a single-polarization SAR image into a waterline detection model, an output label for each pixel is obtained, thus enabling the detection of waterlines on islands.

2. The island waterline detection method based on phase-consistent random walk according to claim 1, characterized in that, The logarithmic Gabor filter is defined as follows: in, and They represent the positions respectively. Image pixels The corresponding angular frequency and direction in polar coordinates and These represent the scale and orientation of the Gabor filter, respectively. and These represent the center frequency and center direction, respectively. Maintain a constant center frequency for different centers. The standard deviation of the Gaussian function in the angular direction; The logarithmic Gabor filter is represented in the spatial domain by the inverse Fourier transform: in, and These represent the current logarithmic Gabor wavelet at scale and direction, respectively. and The real and imaginary parts; the image The response after the spatial filter is expressed as: in, Represents convolution; Indicates the image at position pixel values, express In terms of scale and direction respectively and The actual response, express In terms of scale and direction respectively and The imaginary part of the response.

3. The island waterline detection method based on phase-consistent random walk according to claim 1, characterized in that, Obtaining the phase consistency map specifically includes: Based on image The phase coherence in each direction, as defined by the response of the spatial filter, is as follows: in, , These represent the direction and scale of the log-Gabor filter, respectively. Indicates the interval parameter. Indicates the quantity of scale; For the estimated noise level, It is a constant. This indicates that the included quantity is equal to itself when its value is positive, and zero otherwise; magnitude response. Represented as: This indicates that the log-Gabor filter is at a scale of , direction is Phase deviation function at time, Defined as: in, and They are represented as follows: Weight function Defined as: in, The gain factor is used to control the sharpness of the filter. This is the truncation value. The frequency response extension is expressed as: Where N is the total number of scales considered. It is the amplitude at the nth scale at point x. It is the amplitude at which the maximum response is achieved at x.

4. The island waterline detection method based on phase-consistent random walk according to claim 3, characterized in that, The phase consistency map is transformed from image pixels into a phase consistency metric. : in, A mathematical function representing phase consistency.

5. The island waterline detection method based on phase-consistent random walk according to claim 1, characterized in that, Creating the aforementioned global prior for land and sea specifically includes: Using a phase consistency metric as input, the membership degree between land and sea is obtained through an FCM classifier. This membership degree is... Between, definition For phase consistency The membership degree of a node is defined to reflect the two priors of land and sea. This was then introduced as a priori into the waterline detection model.

6. The island waterline detection method based on phase-consistent random walk according to claim 1, characterized in that, Creating the marked field prior for the impermeable structures connected to the island, specifically including: Assume there exists a line segment detection function. ,get line segment , To locate the bridge, the following conditions must be met: any two line segments must be in opposite directions and have similar absolute values; the distance between the two line segments must be similar; and each line segment must be long enough to mark the bridge. Defined as: in, Represents line segment direction, This represents the calculation of the perpendicular distance between two line segments. This represents the calculation of the length of a line segment. , and These represent the thresholds for angle, distance, and length, respectively.

7. The island waterline detection method based on phase-consistent random walk according to claim 1, characterized in that, The waterline detection model specifically includes: Assuming phase consistency diagram have There are 10 nodes, each representing 10 nodes. One element, while based on the phase consistency graph Generated by superpixels Each node is represented by the mean of each superpixel, and the nodes are integrated. Element nodes and superpixel nodes, totaling A weighted graph is generated using nodes, represented as follows: ,in It is a set of nodes. It is a set of edges. The weight matrix has edge connections; in the node set any node Indicates the first indivual Element nodes or superpixel nodes, in the edge set any edge in Indicates at node In the four-neighbor or eight-neighbor domains of the node The connection, It consists of labeled and unlabeled nodes. This represents a set of labeled nodes. , This indicates that there are K labels, and two category labels are needed for land-sea segmentation. Represents a set of uncategorized labels; Assumption The empty domain of a domain element is ,Depend on The set of spatial domains composed of superpixels of a domain is : in, Indicates the first The spatial set of the n superpixels, the nth Each superpixel The spatial range is represented as .

8. The island waterline detection method based on phase-consistent random walk according to claim 7, characterized in that, Based on the phase consistency diagram Connections between nodes in the domain, connections between nodes in the superpixel domain, and The connection relationships between nodes in the domain and corresponding nodes in the superpixel domain are defined, and the weight matrix is ​​defined. for: Among them, the weight , , , They are defined as follows: in, Phase coherence diagram Nodes in the domain i The value, Phase coherence diagram Nodes in the domain j The value, Indicates the first i The mean value of each superpixel. Indicates the first j The mean of each superpixel; based on Domain elements and superpixels have the same spatial location, and nodes degree Defined as: Introducing cluster membership , will node Prior weights Defined as: in, These are rule parameters, defining the transition probability as: in, and The weights of the two nodes; Based on transition probability From node Walk to the label The probability of node m is: Among them, when hour, ,otherwise ; By setting vectors probability The vector expression is: in, transition probability matrix Defined as: , Let be a vector, where: For any node Calculate the average probability of reaching Its vector representation is: The labels are obtained through calculation: 。

Images