A boundary tracking method for sea ice edge based on sea ice remote sensing

Through the extreme radial projection, cropping and binarization of Arctic sea ice remote sensing data, combined with the noise-resistant expansion-type edge detection operator and sea ice edge point detection method, the problem of low sea ice edge line tracing accuracy and efficiency is solved. , more accurate monitoring and prediction of sea ice edge lines, providing technical support for Arctic shipping and resource development.

CN118411610BActive Publication Date: 2024-11-29PEKING UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410630169.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-21
Publication Date
2024-11-29
Estimated Expiration
2044-05-21

AI Technical Summary

Technical Problem

The existing sea ice edge line tracing methods have problems of insufficient accuracy and low efficiency in Arctic waters, especially in summer, where the ice edge error is large, which affects the navigational analysis of the waterway.

Method used

The edge line tracing method based on sea ice remote sensing is adopted. By using the Arctic sea ice remote sensing data, the polar radial projection, cropping and binarization process is performed, and the noise-resistant expansion-type edge detection operator and sea ice edge point detection method are used to extract the Boundary data of sea ice edge line.

Benefits of technology

It improves the accuracy and efficiency of sea ice edge line boundary tracking, can more accurately monitor and predict changes in Arctic sea ice edge line, and provides decision-making support for the navigation and resource development of Arctic strategic navigation areas.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118411610B_ABST
    Figure CN118411610B_ABST
Patent Text Reader

Abstract

The present disclosure relates to a boundary tracking method for the edge of sea ice based on sea ice remote sensing, which is to perform polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of Arctic sea ice remote sensing data; to clip the polar stereographic projection map based on the spatial coordinate system after polar stereographic projection, and to binarize the clipped polar stereographic projection map to obtain a binary polar stereographic projection map; to perform edge detection on the binary polar stereographic projection map using an anti-noise expansion type edge detection operator to obtain the edge line of the binary polar stereographic projection map; to perform boundary tracking on the edge line of the binary polar stereographic projection map using a sea ice edge point detection method, and to extract the boundary data of the edge line of the binary polar stereographic projection map. The method can solve the problem of insufficient accuracy and low efficiency in boundary tracking and extraction of the edge line of Arctic sea ice, and provide technical support for the monitoring and prediction of sea ice in the Arctic strategic navigation area.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of sea ice monitoring and prediction, and in particular relates to a sea ice edge line boundary tracking method based on sea ice remote sensing. Background Art

[0002] In the past few decades, the Arctic environment has undergone rapid changes. Against the background of global warming, the Arctic ice melt has been accelerating, and the area with no or little ice in summer has expanded and the period has increased, which has made the maritime transportation and military strategic position of the Arctic waters increasingly enhanced. In recent years, the rapid melting of sea ice has provided new navigation options for commercial and scientific research ships passing through the Arctic route, greatly shortening the shipping distance from Asia to Europe. However, the increase in Arctic shipping activities requires a great improvement in the reliability of sea ice edge prediction. Nuclear submarines lurking in the Arctic need to understand the changes in the ice boundary in order to better carry out military operations. In addition, the development of oil, gas and mineral resources in the Arctic waters, as well as Arctic commercial navigation, all require real-time understanding of the real-time changes in the Arctic sea ice edge and its future development. Therefore, tracking and extracting the edge of the sea ice in the Arctic waters is of great significance to ensuring navigation without sea ice obstacles in the Arctic strategic navigation area. Realizing the prediction of short-term ice development can provide decision-making services for military activities, maritime transportation, oil, gas and mineral resource development in the Arctic waters.

[0003] Establishing an accurate Arctic sea ice edge tracking and extraction system has been a major scientific endeavor and faces fundamental challenges. Most of the currently operating sea ice edge tracking systems are based on atmosphere-ice-ocean coupled models. For example, the most advanced sea ice and ocean operational forecast systems include GIOPS (Canadian Global Ice Ocean Prediction System), ACNFS (Arctic Cap Nowcast / Forecast System, the forecast system of the US Navy), and TOPAZ4 (Tunable Optical Profiler for Aerosol and Ozone sea-ice / ocean numerical prediction system, the ocean numerical prediction system of Norway). In these prediction systems, the assimilation methods are very different. For example, both GIOPS and ACNFS systems use three-dimensional variational analysis methods, while TOPAZ4 uses an integrated Kalman filter to assimilate satellite and in situ sea ice and ocean observation data.

[0004] However, due to the chaotic processes in the atmosphere, the predictability of sea ice using physical methods has certain limitations. Considering the accurate identification of the ice-water boundary, it is the key to the analysis of the navigability of the channel. Especially in summer, due to the melting of the ice surface, passive microwave satellite remote sensing tables will underestimate the ice surface area. Therefore, the current sea ice edge line tracking and extraction forecasting models of various methods have large errors in the summer ice edge. From the beginning of the rapid ice growth season (October to December), the ice volume predicted by ACNFS is slower than observed. Averaged over the whole year, the root mean square error of the 6-hour forecast is 79 kilometers, the root mean square error of the 78-hour forecast is 92 kilometers, and the root mean square error of the 126-hour forecast is 104 kilometers.

[0005] Therefore, the current sea ice edge boundary tracking and extraction methods still have problems such as insufficient accuracy and low efficiency. There is an urgent need for a new sea ice edge boundary tracking method based on sea ice remote sensing. Summary of the invention

[0006] The present invention overcomes one of the deficiencies of the prior art and provides a method for tracking the boundary of the sea ice edge based on sea ice remote sensing, which solves the problem of insufficient accuracy and low efficiency in tracking and extracting the boundary of the Arctic sea ice edge, and provides technical support for the monitoring and prediction of sea ice in the Arctic strategic navigation area.

[0007] According to one aspect of the present disclosure, a method for tracking the boundary of a sea ice edge based on sea ice remote sensing is proposed, the method comprising:

[0008] Performing a polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of the Arctic sea ice remote sensing data;

[0009] Clipping the polar stereographic projection image based on the spatial coordinate system after polar stereographic projection to obtain a clipped polar stereographic projection image of the Arctic sea ice remote sensing data;

[0010] Binarization is performed on the clipped polar stereographic projection image of the Arctic sea ice remote sensing data to obtain a binary polar stereographic projection image of the Arctic sea ice remote sensing data;

[0011] The edge detection of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is performed using a noise-resistant expansion type edge detection operator to obtain the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data;

[0012] The edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is tracked by using the sea ice edge point detection method, and the boundary data of the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is extracted.

[0013] In a possible implementation, the clipping of the polar stereographic projection image based on the spatial coordinate system after polar stereographic projection to obtain the clipped polar stereographic projection image of the Arctic sea ice remote sensing data includes:

[0014] Based on the spatial coordinate system after the polar stereographic projection, select the pixel grid point closest to the upper left corner pixel grid point and the lower right corner pixel grid point of the polar stereographic projection image to obtain a first pixel grid point and a second pixel grid point, and record the row number and column number of the first pixel grid point and the second pixel grid point;

[0015] Pixel grid points between the row numbers of the nearest first pixel grid point and the second pixel grid point, and between the column numbers of the first pixel grid point and the second pixel grid point are extracted to obtain a clipped polar stereographic projection map of the Arctic sea ice remote sensing data.

[0016] In a possible implementation, the binarization of the clipped stereographic projection image of the Arctic sea ice remote sensing data includes:

[0017] The non-ice area and data source area values ​​of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data are set to 0, and the sea ice area value of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data is set to 1.

[0018] In a possible implementation, the sea ice edge point detection method is used to track the edge points of the polar stereographic projection image of the binary Arctic sea ice remote sensing data, including:

[0019] Step P1: using a 3*3 tracking template to determine the eight-neighborhood tracking range of the boundary points of the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data;

[0020] Step P2: starting from the upper left corner, scanning the polar stereographic projection image of the binary Arctic sea ice remote sensing data according to a preset order, when scanning the starting boundary pixel point of the edge line, and recording the spatial coordinates and retrieval mark of the starting boundary pixel point; when scanning the ending pixel point of the edge line, outputting all the data of the edge line;

[0021] Step P3: searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point, where the next boundary pixel point is the current boundary pixel point;

[0022] Step P4: When the current boundary pixel point is the end boundary pixel point of the edge line, if the current boundary pixel point returns to the start boundary pixel point or is within the eight-neighborhood tracking range of the start boundary pixel point, the edge line is calibrated as a circular edge line, and the process goes to step P6; otherwise, the edge line is calibrated as a non-closed edge line and the process goes to step P5;

[0023] When the current boundary pixel point is not the end boundary pixel point of the edge line, calibrate the position coordinates of the new boundary pixel point, use the new boundary pixel point as the new start boundary pixel point, and jump to step P3;

[0024] Step P5: If the number of the end boundary pixel points is 1, end the search in the preset direction, jump to step P3, search the eight-neighborhood tracking range of the start boundary pixel point in the reverse direction of the preset direction until another end boundary pixel point of the edge line is found, end the search in the reverse direction of the preset direction, integrate the edge lines in the preset direction and the reverse direction of the preset direction into an edge line searched in the preset direction, and jump to step P6;

[0025] Step P6: taking another ending boundary pixel point of the search in the reverse direction of the preset direction as the starting boundary pixel point, and jumping to step P2 to perform boundary tracing of the polar stereographic projection image of the entire binary Arctic sea ice remote sensing data.

[0026] In a possible implementation, the preset order is in a clockwise direction from left to right and from top to bottom; or in a counterclockwise direction from right to left and from bottom to top.

[0027] In a possible implementation, searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point includes:

[0028] The eight-neighborhood tracking range of the starting boundary pixel point is searched according to the preset direction. When the target boundary pixel point appears for the first time, the target boundary pixel point is used as the next boundary pixel point, and the position code of the target boundary pixel point in the eight-neighborhood tracking range of the starting boundary pixel point is located and recorded.

[0029] In a possible implementation, the calibrating the edge line as a non-closed edge line includes: if there is no new boundary edge point that meets the target condition within the eight-neighborhood tracking range of the current boundary pixel point, calibrating the edge line as a non-closed edge line.

[0030] In a possible implementation, when the current boundary pixel point is not the end boundary pixel point of the edge line, calibrating the position coordinates of a new boundary pixel point and using the new boundary pixel point as a new start boundary pixel point includes:

[0031] If the current boundary pixel point has not returned to the starting boundary pixel point or the eight-neighborhood tracking range of the starting boundary pixel point, and there is a new boundary edge point that meets the conditions within the eight-neighborhood range of the current boundary pixel point, the position code of the new boundary pixel point is recorded, and the new boundary pixel point is used as the new starting boundary pixel point.

[0032] In a possible implementation, the boundaries of the entire binary stereographic projection image of the Arctic sea ice remote sensing data are stored as a one-dimensional coordinate sequence using Freeman coding.

[0033] The disclosed method for tracing the boundary of the sea ice edge line based on sea ice remote sensing comprises the following steps: performing polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of the Arctic sea ice remote sensing data; clipping the polar stereographic projection map based on the spatial coordinate system after the polar stereographic projection to obtain a clipped polar stereographic projection map of the Arctic sea ice remote sensing data; binarizing the clipped polar stereographic projection map of the Arctic sea ice remote sensing data to obtain a binarized polar stereographic projection map of the Arctic sea ice remote sensing data; performing edge detection on the polar stereographic projection map of the binarized Arctic sea ice remote sensing data using an anti-noise expansion type edge detection operator to obtain edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data; and performing boundary tracing on the edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data using a sea ice edge point detection method to extract boundary data of the edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data. It can solve the problems of insufficient accuracy and low efficiency in tracking and extracting the edge of the Arctic sea ice, and provide technical support for the monitoring and prediction of sea ice in the Arctic strategic navigation area.

[0034] Other optional features and technical effects of the embodiments of the present invention are partially described below, and partially can be understood by reading this document. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The elements shown are not limited to the proportions shown in the accompanying drawings. The same or similar reference numerals in the accompanying drawings represent the same or similar elements, wherein:

[0036] Figure 1 A flow chart of a method for tracing the boundary of a sea ice edge line based on sea ice remote sensing according to an embodiment of the present disclosure is shown;

[0037] Figure 2A schematic diagram showing the principle of polar stereographic projection according to an embodiment of the present disclosure is shown;

[0038] Figure 3 A schematic diagram showing the effect of polar stereographic projection clipping and binarization preprocessing of Arctic sea ice data according to an embodiment of the present disclosure is shown;

[0039] Figure 4 A flow chart of a method for tracing the boundary of a sea ice edge line based on sea ice remote sensing according to another embodiment of the present disclosure is shown;

[0040] Figure 5 A schematic diagram of a sea ice edge line search mechanism and tracking direction based on sea ice remote sensing according to an embodiment of the present disclosure is shown;

[0041] Figure 6 A schematic diagram of the termination of annular edge line tracking according to an embodiment of the present disclosure is shown;

[0042] Figure 7 A schematic diagram of terminating non-closed edge line tracking according to an embodiment of the present disclosure is shown;

[0043] Figure 8 A schematic diagram of a Freeman chain code encoding method for image boundary tracking according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific implementation methods and drawings. Here, the exemplary implementation methods of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0045] As used herein, the term "including" and its variations mean open inclusion, i.e., "including but not limited to". Unless otherwise stated, the term "or" means "and / or". The term "based on" means "based at least in part on". The terms "an example embodiment" and "an embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0046] In addition, the steps shown in the flowchart of the accompanying drawings can be executed in a computer such as a set of computer executable instructions. Also, although a logical sequence is shown in the flowchart, in some cases, the steps shown or described can be performed in a sequence different from that here.

[0047] Figure 1 FIG. 1 is a flow chart of a method for tracing the sea ice edge line based on sea ice remote sensing according to an embodiment of the present disclosure. Figure 1As shown, the method may include:

[0048] Step S1: Performing polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of the Arctic sea ice remote sensing data.

[0049] Among them, the latitude and longitude data of Arctic sea ice remote sensing are obtained by satellite equipment and expressed by longitude and latitude (geographic coordinates).

[0050] The principle of polar stereographic projection is to project the points, lines, surfaces and other elements on the earth's surface onto a plane located between the North Pole and the center of the earth, that is, the North Pole is taken as the projection center, and each point on the earth's surface is projected onto this plane along the straight line from the North Pole to the point, which is a coordinate conversion method for projecting data points onto the equatorial plane. Polar stereographic projection mainly reflects the geometric elements (lines, surfaces) of the three-dimensional space of an object on the projection plane for research and processing, providing an image, intuitive, simple and comprehensive quantitative calculation and graphic method, which is widely used in astronomy, navigation, surveying, geography, geological science and other disciplines in structural geology.

[0051] Figure 2 A schematic diagram of the polar stereographic projection principle according to an embodiment of the present disclosure is shown.

[0052] like Figure 2 As shown, the polar stereographic projection has 8 main projection parameters, namely the natural origin latitude φ0, the standard latitude φ F , origin accuracy λ0, scale factor K0 at the natural origin, easting offset F E , Northing offset F N , Assume that the east coordinate of the origin is N E and the assumed north coordinate N of the origin F .

[0053] The formula for converting Arctic sea ice remote sensing data from polar stereographic projection is as follows:

[0054] Formula (1);

[0055] in, Formula (2);

[0056] In the formula, E is east, which refers to the direction of increasing longitude; N is north, which refers to the direction of increasing latitude; θ is the angle of the point north by east, ω=180 0 -θ; φ is the latitude value; a and e are constant parameters.

[0057] The polar stereographic projection conversion formula can be used to convert the longitude and latitude of Arctic sea ice remote sensing data to pixel points on the polar stereographic projection plane, which are represented by two-dimensional coordinates in the polar stereographic projection coordinate system. For example, the projection reference ellipsoid of Arctic sea ice remote sensing data is the WGS-84 ellipsoid, a=6378137.0m, e=0.081819191, the natural origin is the North Pole 0°, the longitude of the natural origin is 0°, the scale factor at the natural origin k0 = 0.9972, and the east and north offsets are both 12288000.0m. This can facilitate the subsequent binarization of polar stereographic projection data and improve the efficiency of boundary tracking of the sea ice edge line.

[0058] Step S2: clipping the polar stereographic projection image based on the spatial coordinate system after polar stereographic projection to obtain a clipped polar stereographic projection image of the Arctic sea ice remote sensing data;

[0059] Step S3: binarizing the clipped polar stereographic projection image of the Arctic sea ice remote sensing data to obtain a binarized polar stereographic projection image of the Arctic sea ice remote sensing data.

[0060] Figure 3 A schematic diagram showing the effect of polar stereographic projection clipping and binarization preprocessing of Arctic sea ice data according to an embodiment of the present disclosure is shown.

[0061] In one example, this step may be based on the spatial coordinate system after the polar stereographic projection, select the pixel grid point closest to the upper left corner pixel grid point and the lower right corner pixel grid point of the polar stereographic projection image, obtain the first pixel grid point and the second pixel grid point, and record the row number and column number of the first pixel grid point and the second pixel grid point;

[0062] The pixel grid points between the row numbers of the nearest first pixel grid point and the second pixel grid point, and between the column numbers of the first pixel grid point and the second pixel grid point are extracted to obtain the clipped polar stereographic projection image of the Arctic sea ice remote sensing data.

[0063] For example, according to the spatial coordinate system after polar stereographic projection, the pixel grid point closest to the upper left pixel grid point (-3850km, 5850) is selected as the first pixel grid point, and the pixel grid point closest to the upper right pixel grid point (3740km, -3750km) is selected as the second pixel grid point. The row number (x-axis of the spatial coordinate system after polar stereographic projection) and column number (y-axis of the spatial coordinate system after polar stereographic projection) of the first pixel grid point and the second pixel grid point are recorded, and the pixel grid points between the row number of the first pixel grid point and the second pixel grid point and between the column number of the first pixel grid point and the second pixel grid point are extracted, and the polar stereographic projection map of the Arctic sea ice remote sensing data of the rectangular grid is obtained by clipping, such as Figure 3As shown. The clipped polar stereographic projection image of the Arctic sea ice remote sensing data is binarized, that is, the non-ice area and data source area values ​​of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data are set to 0 and set as the background color; the sea ice area value of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data is set to 1 and set as the foreground color. The binarized polar stereographic projection image of the Arctic sea ice remote sensing data is shown in Figure 3 As shown, it is convenient to detect edge points and edge lines of polar sea ice remote sensing data, and can avoid interference from irrelevant pixel values, thereby improving the accuracy of detection of edge points and edge lines of polar sea ice remote sensing data.

[0064] Step S4: using a noise-resistant expansion edge detection operator to perform edge detection on the polar stereographic projection image of the binarized Arctic sea ice remote sensing data to obtain edge lines of the polar stereographic projection image of the binarized Arctic sea ice remote sensing data.

[0065] Among them, the noise-resistant expansion edge detection operator is as follows:

[0066] Formula (3);

[0067] Formula (4);

[0068] Among them, is a set symmetric to each other about the origin. It can be understood as using the structural element g(i, j) to detect the original image f(x, y). When the structural element g(i, j) is in the target area of ​​the original image f(x, y) and When there is an intersection, the origin of the target area is included in the expansion set. Expansion is the opposite of erosion, which is to expand the relevant background pixels that are in contact with the target edge and add them to the image set. The final result is to increase the edge pixels of the image.

[0069] Among them, the noise-resistant expansion edge detection operator not only effectively removes the fragmented pixel points of the image, but also ensures the integrity of the edge information of the image; at the same time, the span of the edge points is small, avoiding the influence of topological relationships such as boundary overlap and adjacency on the algorithm. Therefore, the noise-resistant expansion edge detection operator is selected to detect the edge points of the polar projection image of the binarized Arctic sea ice remote sensing data to improve the efficiency of edge point and edge line detection in the polar projection image of the binarized Arctic sea ice remote sensing data.

[0070] Step S5: using a sea ice edge point detection method to track the edge lines of the polar stereographic projection image of the binarized Arctic sea ice remote sensing data, and extracting the edge data of the edge lines of the polar stereographic projection image of the binarized Arctic sea ice remote sensing data.

[0071] Figure 4A flow chart of a method for tracing the boundary of a sea ice edge based on sea ice remote sensing according to another embodiment of the present disclosure is shown.

[0072] In one example, if Figure 4 As shown in FIG. 1 , the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is tracked by using the sea ice edge point detection method, which may include:

[0073] Step P1: using a 3*3 tracking template to determine the eight-neighborhood tracking range of the boundary points of the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data.

[0074] The search range is based on five pixel points in the eight neighborhoods determined by the trend of the line connecting the current boundary point of the edge line and the previous boundary point of the edge line. The boundary of the polar stereographic projection image of the entire binary Arctic sea ice remote sensing data can be stored as a one-dimensional coordinate sequence using Freeman coding.

[0075] Step P2: Starting from the upper left corner, scan the polar stereographic projection image of the binary Arctic sea ice remote sensing data according to a preset order. When the starting boundary pixel point of the edge line is scanned, the spatial coordinates and retrieval mark of the starting boundary pixel point are recorded; when the ending pixel point of the edge line is scanned, all the data of the edge line are output.

[0076] The preset order is a clockwise direction from left to right and from top to bottom; or a counterclockwise direction from right to left and from bottom to top, without any limitation.

[0077] For example, record the starting boundary pixel point P_0 of the edge line, and store the coordinates (x, y) of the starting boundary pixel point P_0 of the edge line in the Freeman chain code of the edge line. The position code of the starting boundary pixel point P_0 of the preset edge line relative to the eight-neighborhood tracking range of the previous edge point is 0 (the position code of the starting boundary pixel point of the non-edge line is not 0), and record its search flag as 1 (set to 0 if it has not been searched, and the search flag is increased by 1 each time it is searched).

[0078] Step P3: searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point, where the next boundary pixel point is the current boundary pixel point.

[0079] Figure 5 A schematic diagram of the sea ice edge line search mechanism and tracking direction based on sea ice remote sensing according to an embodiment of the present disclosure is shown.

[0080] In one example, if Figure 5 As shown, searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point may include:

[0081] The eight-neighborhood tracking range of the starting boundary pixel point is searched according to a preset direction. When the target boundary pixel point appears for the first time, the target boundary pixel point is used as the next boundary pixel point, and the position code of the target boundary pixel point in the eight-neighborhood tracking range of the starting boundary pixel point is located and recorded. That is, the five pixels that meet the corresponding conditions in the eight-neighborhood tracking range of the starting boundary pixel point are searched in a clockwise direction.

[0082] For example, the neighboring pixels of the starting boundary pixel are determined in a clockwise direction. When the target boundary pixel appears for the first time, the target boundary pixel is recorded as the current boundary pixel P_k (k=k+1), and the position code of the target boundary pixel in the eight-neighborhood tracking range of the starting boundary pixel is recorded, and its retrieval flag is set. The next boundary pixel is tracked according to the positional relationship between the target boundary pixel and the starting boundary pixel (the previous boundary pixel).

[0083] After the starting boundary pixel point of the edge line is determined, it is tracked in a certain order within the eight-neighborhood tracking range of the starting boundary pixel point (searched in a clockwise direction) to locate the position of the next boundary pixel point. In order to save time, the search range of the eight-neighborhood tracking range is narrowed.

[0084] like Figure 5 As shown, the current boundary pixel is calibrated as C, the previous edge point of the pixel is set as P, and the eight-neighborhood is divided into two parts by connecting CP and drawing a perpendicular line to CP; then the search range (a part of the divided eight-neighborhood) is selected according to the entry direction of the current boundary pixel, that is, the direction of PC. This method reduces the search range of the edge point by half. C is the current boundary pixel, and P is the previous boundary pixel of C. The black arrow indicates the path from the previous boundary pixel to the current boundary pixel. The gray pixel is the pixel within the tracking range of the eight-neighborhood to be detected, and the arc solid line indicates the tracking direction. Determine whether the eight-neighborhood pixel is the target boundary pixel in a clockwise direction. If a pixel is found to be the target boundary pixel, record the target boundary pixel as the current boundary pixel of the edge line, and continue to track the next boundary pixel based on the positional relationship between the current boundary pixel and the previous boundary pixel. Figure 5 The eight scenarios shown detect each neighboring point in a clockwise direction for edge tracking, which can improve the efficiency of tracking the position of the next boundary pixel point.

[0085] Step P4: When the current boundary pixel point is the ending boundary pixel point of the edge line, if the current boundary pixel point returns to the starting boundary pixel point or the eight-neighborhood tracking range of the starting boundary pixel point, the edge line is calibrated as a circular edge line and the process goes to step P6; otherwise, the edge line is calibrated as a non-closed edge line and the process goes to step P5; wherein, the non-closed edge line is an edge line in which the current boundary pixel point has not returned to the starting boundary pixel point or the eight-neighborhood tracking range of the starting boundary pixel point, and there is no new boundary edge point that meets the target condition within the eight-neighborhood tracking range of the current boundary pixel point.

[0086] When the current boundary pixel point is not the end boundary pixel point of the edge line, the position coordinates of the new boundary pixel point are calibrated, and the new boundary pixel point is used as the new start boundary pixel point, and the process goes to step P3. Wherein, when the current boundary pixel point has not returned to the start boundary pixel point or is within the eight-neighborhood tracking range of the start boundary pixel point, and there is a new boundary edge point that meets the conditions within the eight-neighborhood range of the current boundary pixel point, the position code of the new boundary pixel point is recorded.

[0087] Figure 6 A schematic diagram of the termination of annular edge line tracking according to an embodiment of the present disclosure is shown.

[0088] When tracing the eight-neighborhood of a circular edge line, the possibility that the traced edge line returns to the starting point is relatively small. Figure 6 As shown in the figure, the pixel with pixel ID 1 is the starting pixel of the circular edge line, and the gray pixel is the edge line. The edge line of the target edge is tracked from the starting pixel of the boundary. The black dotted arrow on the way indicates the tracking trend of the edge line. The ordered set of gray pixels along the black dotted arrow is the target edge line to be extracted. This method stipulates that when the edge point is tracked to the eight-neighborhood range of the starting point of the target edge line or just returns to the starting pixel of the boundary, and the entry direction is the same as the edge line tracking trend, the target edge line tracking is stopped, and the circular boundary tracking is completed. Figure 6 For boundary line tracking in the figure, the correct boundary line termination trend should be pointed in the same direction as the solid arrow ① in the figure.

[0089] Step P5: If the number of the ending boundary pixel points is 1, end the search in the preset direction, jump to step P3, search the eight-neighborhood tracking range of the starting boundary pixel point in the reverse direction of the preset direction until another ending boundary pixel point of the edge line is found, end the search in the reverse direction of the preset direction, integrate the edge lines in the preset direction and the reverse direction of the preset direction into an edge line searched in the preset direction, and jump to step P6.

[0090] For example, if there is no boundary pixel within the eight-neighborhood range of the current pixel, the current pixel is considered to be the end boundary pixel, and the number of end boundary pixels is recorded as js_zj=js_zj+1 (the starting value of js_zj is set to 0). If js_zj=1, the starting boundary pixel P_0 is set as the current pixel, and the position code of the starting boundary pixel P_0 is set according to the position relationship between the starting boundary pixel P_0 and the second boundary pixel of the target edge line, and the process jumps to step P3. The target edge line is searched reversely from the starting boundary pixel until another end boundary pixel is found, and the number of end boundary pixels js_zj=js_zj+1 is recorded; If js_zj=2, the edge lines with different tracking directions are reintegrated into an edge line with a clockwise tracking direction, and the process jumps to step P6.

[0091] Figure 7 A schematic diagram of terminating non-closed edge line tracking according to an embodiment of the present disclosure is shown.

[0092] by Figure 7 Take the tracking of the non-closed edge line shown in the figure as an example to illustrate that when the end boundary pixel point of the edge line is tracked (when there is no boundary pixel point within the eight neighborhoods of the target pixel point), this pixel point is set as the end boundary pixel point and the clockwise tracking ends; the start boundary pixel point of the target edge line is returned, and the boundary pixel points are tracked in the counterclockwise direction according to the eight situations shown in the search mechanism until another end boundary pixel point is tracked, and this end boundary pixel point is set as the new start point of the target edge line, and the search of this edge line ends. Figure 7 As shown in the figure, the pixel with pixel ID 1 is the starting boundary pixel of the target edge line, the gray pixel is the boundary line, and the black dotted arrow is the path of boundary tracking. When the boundary pixel with pixel ID 2 is tracked, the clockwise tracking of the current edge line is terminated, and the boundary pixel with pixel ID 1 is returned; the boundary pixel with pixel ID 1 is used as the starting boundary pixel, and the second edge point of the clockwise edge line is set as the previous edge point of the current edge starting point, and the next boundary pixel is tracked in the counterclockwise direction to search for the route. Figure 7 The solid arrow in the middle of the edge line is used to track the edge line until the boundary pixel point with pixel identification 3 (the end boundary pixel point of the current edge line) is tracked, and the edge line tracking is stopped. The pixel point with pixel identification 3 is set as the new starting boundary pixel point of this edge line, and the tracking of this edge line is finally ended.

[0093] Step P6: taking another ending boundary pixel point of the search in the reverse direction of the preset direction as the starting boundary pixel point, and jumping to step P2 to perform boundary tracing of the polar stereographic projection image of the entire binary Arctic sea ice remote sensing data.

[0094] Figure 8 A schematic diagram of a Freeman chain code encoding method for image boundary tracking according to an embodiment of the present disclosure is shown.

[0095] For the boundary tracing of the polar stereographic projection image of the entire binary Arctic sea ice remote sensing data, the Freeman chain code (i.e., coordinate code) in the eight-neighborhood direction is selected, and the pixel points along the edge of the image are tracked in a clockwise direction from the starting boundary pixel point of the target edge line. The Freeman chain code is used to store the positional relationship between the pixels and the trend of the edge line until returning to the starting boundary pixel point of the edge line or the ending boundary pixel point of the edge line.

[0096] like Figure 8 As shown in the figure, Freeman chain coding is used to describe the edge line of the area. Gray pixels represent the boundary pixels of the edge line, demonstrating the way of recording boundary pixels by Freeman chain code under the eight-connected relationship. The coordinates of the starting boundary pixel of the edge line are (x, y), and the Freeman chain code recording its edge line is expressed as (x, y): 077766665644444442221132101. Among them, the eight direction symbols defining the pixel points within the eight-neighborhood tracking range are: 0, 1, 2, 3, 4, 5, 6, 7. The starting boundary pixel point of the edge line plus a set of sequence consisting of these direction symbols is called Freeman chain code. Freeman chain code is a common technology in this field and will not be described in detail here.

[0097] The disclosed method for tracing the boundary of the sea ice edge line based on sea ice remote sensing comprises the following steps: performing polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of the Arctic sea ice remote sensing data; clipping the polar stereographic projection map based on the spatial coordinate system after the polar stereographic projection to obtain a clipped polar stereographic projection map of the Arctic sea ice remote sensing data; binarizing the clipped polar stereographic projection map of the Arctic sea ice remote sensing data to obtain a binarized polar stereographic projection map of the Arctic sea ice remote sensing data; performing edge detection on the polar stereographic projection map of the binarized Arctic sea ice remote sensing data using an anti-noise expansion type edge detection operator to obtain edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data; and performing boundary tracing on the edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data using a sea ice edge point detection method to extract boundary data of the edge lines of the polar stereographic projection map of the binarized Arctic sea ice remote sensing data. It can detect sea ice edge points and track and extract sea ice edge lines from sea ice binary data, solving the problems of insufficient accuracy and low efficiency in tracking and extracting Arctic sea ice edge lines, and providing technical support for monitoring and predicting sea ice in the Arctic strategic navigation area.

[0098] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A method for tracking the edge of sea ice based on sea ice remote sensing, characterized in that: The method comprises: Performing a polar stereographic projection on the longitude and latitude data of Arctic sea ice remote sensing to obtain a polar stereographic projection map of the Arctic sea ice remote sensing data; Clipping the polar stereographic projection image based on the spatial coordinate system after polar stereographic projection to obtain the clipped polar stereographic projection image of the Arctic sea ice remote sensing data, which includes selecting the nearest pixel grid point to the upper left pixel grid point and the lower right pixel grid point of the polar stereographic projection image based on the spatial coordinate system after polar stereographic projection to obtain a first pixel grid point and a second pixel grid point, and recording the row number and column number of the first pixel grid point and the second pixel grid point; Extracting pixel grid points between the row numbers of the nearest first pixel grid point and the second pixel grid point, and between the column numbers of the first pixel grid point and the second pixel grid point, to obtain a clipped polar stereographic projection image of the Arctic sea ice remote sensing data; Binarization is performed on the clipped polar stereographic projection image of the Arctic sea ice remote sensing data to obtain a binary polar stereographic projection image of the Arctic sea ice remote sensing data; The edge detection of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is performed using a noise-resistant expansion type edge detection operator to obtain the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data; The edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is tracked by using the sea ice edge point detection method, and the edge data of the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data is extracted; The method of using the sea ice edge point detection method to track the edge points of the polar stereographic projection image of the binary Arctic sea ice remote sensing data includes: Step P1: using a 3*3 tracking template to determine the eight-neighborhood tracking range of the boundary points of the edge line of the polar stereographic projection image of the binary Arctic sea ice remote sensing data; Step P2: starting from the upper left corner, scanning the polar stereographic projection image of the binary Arctic sea ice remote sensing data according to a preset order, when scanning the starting boundary pixel point of the edge line, and recording the spatial coordinates and retrieval mark of the starting boundary pixel point; when scanning the ending pixel point of the edge line, outputting all the data of the edge line; Step P3: searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point, where the next boundary pixel point is the current boundary pixel point; Step P4: When the current boundary pixel point is the end boundary pixel point of the edge line, if the current boundary pixel point returns to the start boundary pixel point or is within the eight-neighborhood tracking range of the start boundary pixel point, the edge line is calibrated as a circular edge line, and the process goes to step P6; otherwise, the edge line is calibrated as a non-closed edge line and the process goes to step P5; When the current boundary pixel point is not the end boundary pixel point of the edge line, calibrate the position coordinates of the new boundary pixel point, use the new boundary pixel point as the new start boundary pixel point, and jump to step P3; Step P5: If the number of the end boundary pixel points is 1, end the search in the preset direction, jump to step P3, search the eight-neighborhood tracking range of the start boundary pixel point in the reverse direction of the preset direction until another end boundary pixel point of the edge line is found, end the search in the reverse direction of the preset direction, integrate the edge lines in the preset direction and the reverse direction of the preset direction into an edge line searched in the preset direction, and jump to step P6; Step P6: taking another ending boundary pixel point of the search in the reverse direction of the preset direction as the starting boundary pixel point, and jumping to step P2 to perform boundary tracing of the polar stereographic projection image of the entire binary Arctic sea ice remote sensing data.

2. The boundary tracking method according to claim 1, characterized in that: The binarization of the clipped stereographic projection image of the Arctic sea ice remote sensing data includes: The non-ice area and data source area values ​​of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data are set to 0, and the sea ice area value of the clipped polar stereographic projection image of the Arctic sea ice remote sensing data is set to 1.

3. The boundary tracking method according to claim 1, characterized in that: The preset order is a clockwise direction from left to right and from top to bottom; or a counterclockwise direction from right to left and from bottom to top.

4. The boundary tracking method according to claim 1, characterized in that: The step of searching the eight-neighborhood tracking range of the starting boundary pixel point according to a preset direction to locate the position of the next boundary pixel point includes: The eight-neighborhood tracking range of the starting boundary pixel point is searched according to the preset direction. When the target boundary pixel point appears for the first time, the target boundary pixel point is used as the next boundary pixel point, and the position code of the target boundary pixel point in the eight-neighborhood tracking range of the starting boundary pixel point is located and recorded.

5. The boundary tracking method according to claim 4, characterized in that: The step of marking the edge line as a non-closed edge line includes: If there is no new boundary edge point that meets the target condition within the eight-neighborhood tracking range of the current boundary pixel point, the edge line is marked as a non-closed edge line.

6. The boundary tracking method according to claim 5, characterized in that: When the current boundary pixel point is not the end boundary pixel point of the edge line, calibrating the position coordinates of a new boundary pixel point and using the new boundary pixel point as a new start boundary pixel point includes: If the current boundary pixel point has not returned to the starting boundary pixel point or the eight-neighborhood tracking range of the starting boundary pixel point, and there is a new boundary edge point that meets the conditions within the eight-neighborhood range of the current boundary pixel point, the position code of the new boundary pixel point is recorded, and the new boundary pixel point is used as the new starting boundary pixel point.

7. The boundary tracking method according to claim 1, characterized in that: The boundaries of the entire binary polar stereographic projection image of the Arctic sea ice remote sensing data are stored as a one-dimensional coordinate sequence using Freeman coding.

Citation Information

Patent Citations

  • Sea ice cube watershed segmentation method based on pixel optimization

    CN111862117A

  • Surface feature-based yellow river delta turbid sea area sea ice classification method

    CN113269026A