A sea ice segmentation method and device based on convex optimization and a medium

Abstract

The application discloses a sea ice segmentation method and device based on convex optimization and a medium, and relates to the technical field of image processing.The method comprises the following steps: performing ice-water classification on a sea ice image of a sea area to be segmented to obtain a sea ice image binary classification image; performing preliminary sea ice segmentation by using a traditional watershed algorithm to obtain a plurality of initial sea ice segmentation lines; determining the initial convexity of each initial sea ice based on the connected components of each initial sea ice, and determining initial non-independent sea ice based on the initial sea ice segmentation lines and the initial convexity of each initial sea ice; and performing morphological erosion on the connected components of the initial non-independent sea ice for multiple iterations by using a morphological erosion method based on Euclidean distance transformation, so as to obtain a target sea ice segmentation result of the sea area to be segmented.The application reduces the over-segmentation phenomenon in the segmentation of long-strip sea ice, and improves the accuracy and reliability of sea ice segmentation.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a sea ice segmentation method, apparatus and medium based on convex optimization. Background Technology

[0002] Sea ice is a crucial component of polar and high-latitude seas, exerting a wide-ranging influence on the global climate system, marine ecosystems, and human socio-economic activities through its unique physical properties and extensive geographical distribution. Sea ice not only regulates ocean-atmosphere energy exchange, playing a vital role in global climate stability, but also provides habitats and breeding grounds for numerous organisms, maintaining ecosystem balance. However, global warming has led to a reduction and thinning of sea ice compared to the past, making it more susceptible to weather-related melting and fragmentation, exacerbating sea-level rise and threatening coastal ecosystems and residential safety. Furthermore, the cutting, collision, and entrapment of broken sea ice by drifting ice can affect maritime traffic safety, increasing navigation difficulty and accident risks, posing challenges to sea ice resource development, scientific research, and other activities.

[0003] Sea ice fragmentation size distribution is a key parameter in polar and high-latitude environments, influencing the evolution of sea ice dynamics and thermodynamic processes. When sea ice breaks up under external forces, it consists of both large ice blocks and smaller fragments that break off from larger blocks. Large ice blocks can alter the local temperature and salinity distribution, while smaller fragments are more susceptible to dynamic factors such as wind, waves, tides, and currents, affecting ocean dynamic processes. Sea ice fragmentation is a crucial process in studying sea ice fragmentation size distribution; the accuracy of its results is essential for calculating ocean momentum and heat balance, as well as predicting marine environmental conditions.

[0004] Previous studies have proposed numerous sea ice segmentation methods that have effectively advanced the identification of sea ice fragment size distribution. In particular, traditional sea ice methods have automated the sea ice segmentation process. Currently, traditional methods are widely used in sea ice segmentation of ice-covered images. These methods consume relatively few computational resources and preserve the original size and shape of the sea ice to the greatest extent possible. However, they still exhibit over-segmentation issues for elongated sea ice. Summary of the Invention

[0005] The purpose of this application is to provide a sea ice segmentation method, apparatus, and medium based on convex optimization, so as to reduce over-segmentation in the segmentation of long strip sea ice and improve the accuracy and reliability of sea ice segmentation.

[0006] To achieve the above objectives, this application provides the following solution.

[0007] Firstly, this application provides a sea ice segmentation method based on convex optimization, comprising:

[0008] Acquire sea ice images of the sea area to be segmented;

[0009] Based on the sea ice image, ice and water classification is performed to obtain a binarized classification image of the sea ice image and the connected components of multiple initial sea ices in the binarized classification image of the sea ice image;

[0010] Using the traditional watershed algorithm, preliminary sea ice segmentation is performed based on the binarized classification image of the sea ice image and the connected components of each initial sea ice, resulting in multiple initial sea ice segmentation lines;

[0011] The initial convexity of each initial sea ice is determined based on the connected components of each initial sea ice, and the initial non-independent sea ice is determined based on the dividing lines of each initial sea ice and the initial convexity of each initial sea ice.

[0012] A morphological erosion method based on Euclidean distance transformation is used to perform multiple iterations of morphological erosion on the connected components of the initial non-independent sea ice, thereby obtaining the target sea ice segmentation result of the sea area to be segmented.

[0013] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described convex optimization-based sea ice segmentation method.

[0014] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned sea ice segmentation method based on convex optimization.

[0015] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0016] This application discloses a sea ice segmentation method, apparatus, and medium based on convex optimization. First, ice-water classification is performed based on sea ice images to obtain binary classification images of the sea ice. Then, using the traditional watershed algorithm, preliminary sea ice segmentation is performed based on the binary classification images and the connected components of each initial sea ice segment, resulting in multiple initial sea ice segmentation lines. Second, the initial convexity of each initial sea ice segment is determined based on its connected components, and initial non-independent sea ice segments are determined based on the initial sea ice segmentation lines and their initial convexities. Finally, a morphological erosion method based on Euclidean distance transform is used to perform multiple iterative morphological erosions on the connected components of the initial non-independent sea ice segments, thereby obtaining the target sea ice segmentation result for the sea area to be segmented. This application, based on the traditional watershed algorithm, introduces convexity for multiple iterative morphological erosions, reducing over-segmentation in long strip sea ice segmentation and improving the accuracy and reliability of sea ice segmentation. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of a sea ice segmentation method based on convex optimization provided in an embodiment of this application.

[0019] Figure 2 This is a diagram of the architecture of a sea ice segmentation method based on convex optimization.

[0020] Figure 3 Images of sea ice acquired by an optical satellite.

[0021] Figure 4 Binarization and classification of sea ice images.

[0022] Figure 5 It is a grayscale image.

[0023] Figure 6 This is a schematic diagram of the convexity of sea ice.

[0024] Figure 7 This is a map showing the updated sea ice segmentation results after a certain number of iterations.

[0025] Figure 8 The result of segmenting the target sea ice.

[0026] Figure 9 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0027] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0028] The purpose of this application is to provide a sea ice segmentation method, apparatus, and medium based on convex optimization, which aims to reduce over-segmentation in the segmentation of long strip sea ice and improve the accuracy and reliability of sea ice segmentation.

[0029] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] In one exemplary embodiment, such as Figure 1 and Figure 2 As shown, a sea ice segmentation method based on convex optimization is provided, which includes the following steps.

[0031] Step 1: Obtain sea ice images of the sea area to be segmented.

[0032] Specifically, sea ice images of the area to be segmented, acquired using optical satellites, have a spatial resolution of 1 meter, such as... Figure 3 As shown.

[0033] Step 2: Perform ice-water classification based on sea ice images to obtain a binarized classification image of the sea ice image and the connected components of multiple initial sea ices in the binarized classification image of the sea ice image.

[0034] As an optional implementation, step 2 includes the following steps.

[0035] Step 21: Input the sea ice image to be segmented into the trained random forest model to obtain a binary classification image of the sea ice image.

[0036] Specifically, the binarized classification images of sea ice corresponding to the sea ice images acquired by optical satellites are as follows: Figure 4 As shown.

[0037] Step 22: Determine the connected components of multiple initial sea ices based on the binarized classification image of the sea ice image.

[0038] Step 3: Using the traditional watershed algorithm, preliminary sea ice segmentation is performed based on the binarized classification image of the sea ice image and the connected components of each initial sea ice segment to obtain multiple initial sea ice segmentation lines.

[0039] As an optional implementation, step 3 includes the following steps.

[0040] Step 31: Identify any initial sea ice as the sea ice to be segmented.

[0041] Step 32: Determine the initial nearest Euclidean distance of each pixel in the connected component of the sea ice to be segmented; the initial nearest Euclidean distance is the minimum Euclidean distance from each pixel in the connected component to the boundary of the seawater in the binarized classification image of the sea ice image.

[0042] Specifically, the formula for calculating the initial nearest Euclidean distance of any pixel in the connected components of the sea ice to be segmented is as follows:

[0043] .

[0044] in, For the connected components of the sea ice to be segmented The initial nearest Euclidean distance of the pixel at that location; The coordinates of the pixels representing the boundaries of seawater in the binarized classification image of the sea ice image; To classify the seawater boundary in a binary classification image of sea ice.

[0045] Step 33: Based on the initial nearest Euclidean distance of each pixel in the connected components of the sea ice to be segmented, determine the grayscale value of each pixel in the connected components of the sea ice to be segmented, thereby obtaining a grayscale image of the sea area to be segmented. The grayscale image is as follows: Figure 5 As shown.

[0046] Specifically, the formula for calculating the grayscale value of any pixel in the connected components of the sea ice to be segmented is as follows:

[0047] .

[0048] in, For the connected components of the sea ice to be segmented The grayscale value of the pixel at that location; The minimum initial nearest Euclidean distance among all pixels in the connected components of the sea ice to be segmented; This represents the maximum initial nearest Euclidean distance among all pixels in the connected components of the sea ice to be segmented.

[0049] Step 34: Based on the grayscale values ​​of each pixel in the grayscale image, divide each pixel into multiple lowest points of the catchment basin and multiple lowest points of the non-catchment basin.

[0050] Specifically, step 34 includes the following steps.

[0051] Step 341: Determine any pixel in the grayscale image as the point to be judged, and determine all pixels in the 8-neighborhood of the point to be judged as auxiliary points.

[0052] Step 342: If the gray value of the point to be judged is less than the gray values ​​of all the corresponding auxiliary points, then the point to be judged is determined as the lowest point of the catchment basin.

[0053] Step 343: Determine all pixels in the grayscale image except for the lowest point of the catchment basin as the lowest point of the non-catchment basin.

[0054] Step 35: Determine the gray value of each pixel in the grayscale image as the initial water level height of the corresponding pixel.

[0055] Step 36: Based on the initial water level of all pixels in the sea area to be segmented, update the water level of the lowest point of each non-catchment basin in multiple rounds to obtain the target water level of the lowest point of each non-catchment basin.

[0056] As an optional implementation, step 36 includes the following steps.

[0057] Step 361: Determine the lowest point of any non-catchment basin as the point to be updated, and determine the pixels in the 8-neighborhood of the point to be updated as neighborhood points.

[0058] Step 362: Determine any round as the current round.

[0059] Step 363: Based on the initial water level of the point to be updated and the water level of each neighboring point in the previous round, determine the water level of the point to be updated in the current round; when the current round is the initial round, the water level of each neighboring point in the previous round is the initial water level of each neighboring point; when the current round is not the initial round and the neighboring point is the lowest point of the catchment basin, the water level of the neighboring point in the previous round is the initial water level of the neighboring point.

[0060] Specifically, the lowest point of any non-catchment basin is at the 1st The formula for calculating the water level height for each cycle is:

[0061] .

[0062] in, for The lowest point of the non-catchment basin is at the Water level height in each round; for The set of neighborhood points of the lowest point in the non-catchment basin; The first of the set of neighborhood points of the lowest point in a non-catchment basin The neighboring points at the th ... Water level height in each round, ; for The initial water level height at the lowest point of the non-catchment basin.

[0063] Step 364: Determine whether the update stop condition is met; the update stop condition is whether the difference between the water level of the lowest point of each non-catchment basin in the current cycle and the water level in the previous cycle is less than the first preset difference value.

[0064] Step 365: If the update stop condition is met, then the water level of the lowest point in the non-catchment basin in the current round is determined as the target water level of the corresponding lowest point in the non-catchment basin.

[0065] Step 366: If the update stop condition is not met, proceed to the next round of updates until the update stop condition is met, and obtain the target water level height of the lowest point of each non-catchment basin.

[0066] Step 37: Determine the initial water level of the lowest point of each catchment basin as the target water level of the corresponding lowest point of the catchment basin.

[0067] Step 38: Based on the target water level height of all pixels in the sea area to be segmented, determine multiple initial sea ice segmentation lines.

[0068] As an optional implementation, step 38 includes the following steps.

[0069] Step 381: Determine whether the difference between the target water level heights of any two pixels in the sea area to be segmented is less than the second preset difference value.

[0070] Step 382: If the difference is less than the second preset value, then one of the two pixels is determined as the dividing point.

[0071] Step 383: Determine multiple initial sea ice dividing lines based on each dividing point.

[0072] Step 4: Determine the initial convexity of each initial sea ice based on the connected components of each initial sea ice, and determine the initial non-independent sea ice based on the dividing lines of each initial sea ice and the initial convexity of each initial sea ice.

[0073] As an optional implementation, step 4 includes the following steps.

[0074] Step 41: Determine the initial convexity of each initial sea ice based on the perimeter of the convex hull of the connected component of each initial sea ice and the perimeter of the contour of the connected component of each initial sea ice.

[0075] As an optional implementation method, such as Figure 6 As shown, the formula for calculating the convexity of any sea ice is:

[0076] .

[0077] in, The convexity of sea ice; Let be the perimeter of the convex hull of the connected components of the sea ice; Let be the perimeter of the outline of the connected components of the sea ice.

[0078] Step 42: Determine all initial sea ice that meets the non-independent condition as initial non-independent sea ice; the non-independent condition is that there is no sea ice dividing line in the connected component and the convexity is greater than the convexity threshold.

[0079] Step 5: Using a morphological erosion method based on Euclidean distance transformation, perform multiple iterative morphological erosions on the connected components of the initial non-independent sea ice to obtain the target sea ice segmentation result of the sea area to be segmented.

[0080] As an optional implementation, step 5 includes the following steps.

[0081] Step 501: Determine any iteration number as the current iteration number.

[0082] Step 502: Determine all updated non-independent sea ice segments from the previous iteration as non-independent sea ice segments in the current iteration; determine all updated seawater boundaries from the previous iteration as seawater boundaries in the current iteration; and determine the updated sea ice segmentation lines from the previous iteration as sea ice segmentation lines in the current iteration. When the current iteration is the initial iteration, the updated non-independent sea ice segments from the previous iteration are the initial non-independent sea ice segments, the updated seawater boundaries from the previous iteration are the seawater boundaries in the binary classification image of the sea ice image, and the updated sea ice segmentation lines from the previous iteration are the initial sea ice segmentation lines. The sea ice segmentation result image after obtaining the updated sea ice segmentation lines at a certain iteration is shown below. Figure 7 As shown, the red lines represent the sea ice demarcation lines deleted at this iteration number.

[0083] Step 503: Determine the nearest Euclidean distance of each pixel in the connected component of the non-independent sea ice at the current iteration number; the nearest Euclidean distance is the minimum Euclidean distance from each pixel in the connected component of the non-independent sea ice to the boundary of the seawater.

[0084] Specifically, the formula for calculating the nearest Euclidean distance of each pixel in the connected component of the non-independent sea ice at the current iteration number is as follows:

[0085] .

[0086] in, For the connected components of non-independent sea ice at the current iteration number The nearest Euclidean distance of the pixel at the current iteration number; These are the pixel coordinates of the seawater boundary at the current iteration number; This represents the boundary of the seawater at the current iteration number.

[0087] Step 504: Determine whether the erosion condition is met; the erosion condition is that the nearest Euclidean distance of the pixels in the connected components of the non-independent sea ice at the current iteration number is greater than the distance threshold at the current iteration number.

[0088] Step 505: Erode the pixels in the connected components of the non-independent sea ice at the current iteration number that do not meet the erosion conditions into seawater, thereby obtaining the eroded connected components of the non-independent sea ice at the current iteration number.

[0089] Specifically, pixels in the connected components of non-independent sea ice at the current iteration number that meet the corrosion conditions remain as seawater pixels.

[0090] Step 506: Determine the updated seawater boundary for the current iteration number based on the eroded connected components of the non-independent sea ice at the current iteration number.

[0091] Step 507: Delete the sea ice segmentation lines in the current iteration that coincide with the eroded connected components of the non-independent sea ice in the current iteration, and obtain the updated sea ice segmentation lines in the current iteration.

[0092] Step 508: Determine the convexity of each non-independent sea ice in the current iteration number based on the eroded connected components of each non-independent sea ice in the current iteration number.

[0093] Specifically, step 508 includes: determining the convexity of each non-independent sea ice in the current iteration number based on the perimeter of the convex hull of the eroded connected component of each non-independent sea ice in the current iteration number and the perimeter of the contour of the eroded connected component of each non-independent sea ice in the current iteration number.

[0094] Step 509: Based on the updated sea ice segmentation line at the current iteration number and the convexity of each non-independent sea ice at the current iteration number, determine the updated non-independent sea ice at the current iteration number.

[0095] Specifically, step 509 includes: determining the convexity of each non-independent sea ice in the current iteration number that satisfies the non-independent condition as the updated non-independent sea ice in the current iteration number; the non-independent condition is that there is no sea ice dividing line in the connected component and the convexity is greater than the convexity threshold.

[0096] Step 510: Determine whether the corrosion stopping condition is met; the corrosion stopping condition is that the preset number of iterations is reached or the convexity of each non-independent sea ice at the current iteration number is less than or equal to the convexity threshold.

[0097] Step 511: If the erosion stopping condition is met, the eroded connected components of the non-independent sea ice at the current iteration number are determined as the target sea ice segmentation result for the sea area to be segmented.

[0098] Step 512: If the corrosion stopping condition is not met, proceed with the next iteration of corrosion until the corrosion stopping condition is met, and obtain the target sea ice segmentation result for the sea area to be segmented.

[0099] Specifically, the target sea ice segmentation results are as follows: Figure 8 As shown.

[0100] In one exemplary embodiment, a computer device is provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer program to implement a convex optimization-based sea ice segmentation method.

[0101] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements a convex optimization-based sea ice segmentation method.

[0102] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a sea ice segmentation method based on convex optimization.

[0103] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0104] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0105] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0106] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0107] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0108] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A sea ice segmentation method based on convex optimization, characterized in that, The sea ice segmentation method based on convex optimization comprises: An ice-water classification is performed based on the sea ice image to obtain a sea ice image binary classification image and a plurality of initial sea ice connected components in the sea ice image binary classification image. An ice-water classification is performed based on the sea ice image to obtain a sea ice image binary classification image and a plurality of initial sea ice connected components in the sea ice image binary classification image. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The connected component of the initial non-independent sea ice is iteratively morphologically eroded using a morphological erosion method based on Euclidean distance transformation, so as to obtain a target sea ice segmentation result of the sea area to be segmented.

2. The convex optimization-based sea ice segmentation method of claim 1, wherein, An ice-water classification is performed based on the sea ice image to obtain a sea ice image binary classification image and a plurality of initial sea ice connected components in the sea ice image binary classification image. An ice-water classification is performed based on the sea ice image to obtain a sea ice image binary classification image and a plurality of initial sea ice connected components in the sea ice image binary classification image. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line.

3. The convex optimization based sea ice segmentation method of claim 1, wherein, The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line.

4. The convex optimization-based sea ice segmentation method of claim 3, wherein, The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on the initial convexity of each initial sea ice and the initial sea ice segmentation line. The initial convexity of each initial sea ice is determined based on the connected component of each initial sea ice, and the initial non-independent sea ice is determined based on determining the water level height of the to-be-updated point in the current round based on the initial water level height of the to-be-updated point and the water level height of each neighboring point in the last round; when the current round is the initial round, the water level height of each neighboring point in the last round is the initial water level height of each neighboring point; when the current round is a non-initial round and the neighboring point is the lowest point of the catchment basin, the water level height of the neighboring point in the last round is the initial water level height of the neighboring point; determining whether the update stop condition is met; the update stop condition is that a preset round is reached or whether the difference between the water level height of each non-catchment-basin lowest point in the current round and the water level height in the last round is less than a first preset difference value; if the update stop condition is met, determining the water level height of the non-catchment-basin lowest point in the current round as the target water level height of the corresponding non-catchment-basin lowest point; if the update stop condition is not met, performing the update in the next round until the update stop condition is met, and obtaining the target water level height of each non-catchment-basin lowest point.

5. The convex optimization based sea ice segmentation method of claim 3, wherein, determining a plurality of initial sea ice segmentation lines based on the target water level height of all pixel points in the to-be-segmented sea area, comprising: determining whether the difference between the target water level heights of any two pixel points in the to-be-segmented sea area is less than a second preset difference value; if the difference is less than the second preset difference value, determining one of the two pixel points as a segmentation point; determining a plurality of initial sea ice segmentation lines based on each segmentation point.

6. The convex optimization-based sea ice segmentation method of claim 1, wherein, determining the initial convexity of each initial sea ice based on each connected component of the initial sea ice, and determining initial non-independent sea ice based on the initial sea ice segmentation lines and the initial convexity of each initial sea ice, comprising: determining the initial convexity of each initial sea ice according to the perimeter of the convex hull of each connected component of the initial sea ice and the perimeter of the outline of each connected component of the initial sea ice; determining each initial sea ice that meets the non-independent condition as an initial non-independent sea ice; the non-independent condition is that there is no sea ice segmentation line in the connected component and the convexity is greater than a convexity threshold.

7. The convex optimization-based sea ice segmentation method of claim 1, wherein, obtaining the target sea ice segmentation result of the to-be-segmented sea area by using a morphological erosion method based on Euclidean distance transformation to perform morphological erosion on the connected component of the initial non-independent sea ice for multiple iterations, comprising: determining any iteration number as the current iteration number; determining each non-independent sea ice in the last iteration as the non-independent sea ice in the current iteration, determining the boundary of the seawater in the last iteration as the boundary of the seawater in the current iteration, and determining the sea ice segmentation line in the last iteration as the sea ice segmentation line in the current iteration; when the current iteration number is the initial iteration number, each non-independent sea ice in the last iteration is the initial non-independent sea ice, the boundary of the seawater in the last iteration is the boundary of the seawater in the sea ice binary classification image, and the sea ice segmentation line in the last iteration is the initial sea ice segmentation line; determining the nearest Euclidean distance of each pixel point in the connected component of the non-independent sea ice in the current iteration; the nearest Euclidean distance is the minimum value of the Euclidean distance from the pixel point in the connected component of the non-independent sea ice to each pixel of the boundary of the seawater. determining whether a corrosion condition is satisfied; the corrosion condition is that a pixel in a connected component of the non-independent sea ice at the current iteration number has a nearest Euclidean distance greater than a distance threshold at the current iteration number; eroding the pixel in the connected component of the non-independent sea ice at the current iteration number that does not satisfy the corrosion condition to sea water, thereby obtaining an eroded connected component of the non-independent sea ice at the current iteration number; determining an updated boundary of the sea water at the current iteration number according to the eroded connected component of the non-independent sea ice at the current iteration number; deleting the sea ice segmentation line at the current iteration number that overlaps with the eroded connected component of the non-independent sea ice at the current iteration number, to obtain an updated sea ice segmentation line at the current iteration number; determining a convexity of each non-independent sea ice at the current iteration number based on the eroded connected component of each non-independent sea ice at the current iteration number, respectively; determining updated non-independent sea ice at the current iteration number based on the updated sea ice segmentation line at the current iteration number and the convexity of each non-independent sea ice at the current iteration number; determining whether a corrosion stop condition is satisfied; the corrosion stop condition is that a preset iteration number is reached or the convexity of each non-independent sea ice at the current iteration number is less than or equal to a convexity threshold; if the corrosion stop condition is satisfied, determining the eroded connected component of the non-independent sea ice at the current iteration number as a target sea ice segmentation result of the sea area to be segmented; if the corrosion stop condition is not satisfied, performing corrosion at a next iteration number until the corrosion stop condition is satisfied, to obtain the target sea ice segmentation result of the sea area to be segmented.

8. The convex optimization-based sea ice segmentation method of claim 6, wherein, The formula for calculating the convexity of any sea ice is: ； wherein, is a convexity of the sea ice; is a perimeter of a convex hull of a connected component of the sea ice; is a perimeter of a contour of a connected component of the sea ice.

9. A computer apparatus comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the sea ice segmentation method based on convex optimization of any one of claims 1-8.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the sea ice segmentation method based on convex optimization of any one of claims 1-8.

Images