Track image-oriented ship route extraction method and storage medium

By combining 3D data extraction and bilinear interpolation techniques with edge detection and Mercator projection, the subjective dependence and complexity of existing ship route extraction methods are resolved, achieving efficient and automated route extraction and analysis, and improving the accuracy and processing efficiency of route information.

CN119006373APending Publication Date: 2024-11-22SHANGHAI SHIP & SHIPPING RES INST CO LTD +1
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410999025.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

Existing methods for extracting ship routes rely on the experience of navigators and cannot make full use of big data and ocean forecast information. Methods based on traditional oceanography have limitations when dealing with variable and nonlinear ocean data. Methods based on mathematical models are complex and require high parameter calibration. Methods based on two-dimensional image processing cannot accurately determine the route boundary and increase the error. Machine learning methods are complex and computationally difficult.

Method used

A3D data extraction methods are used to transform AIS data into a 3D route density scatter plot. Elevation data is identified through bilinear interpolation and latitude interpolation and converted into a 2D latitude and longitude coordinate system. Combined with edge detection and smoothing techniques, route reconstruction and smoothing are achieved. Mercator projection is used for coordinate transformation, and route links are established by associating pixels through a set distance threshold.

Benefits of technology

It achieves high-precision and automated ship route extraction, overcomes the influence of subjective judgment, saves human resources, improves work efficiency, can accurately determine route boundaries and reduce errors, has a simple processing process, low computational difficulty, and is suitable for large-scale ship trajectory data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119006373A_ABST
    Figure CN119006373A_ABST
Patent Text Reader

Abstract

The invention discloses a ship route extraction method for a track image and a storage medium, which can automatically extract a large amount of ship route information without being influenced by subjective factors, completely overcome the influence of noise, can perform high-precision identification and extraction on ship routes with different busy degrees, does not need a complex mathematical model and selection with high requirements on parameters, and is high in efficiency. Manpower resources are saved, and the working efficiency is improved. According to the technical scheme, the method is characterized by comprising the following steps that S1, a three-dimensional data extraction method is adopted for a ship course thermodynamic diagram image formed by ship AIS data, and the ship course thermodynamic diagram image is converted into a three-dimensional ship course density scatter image; s2, carrying out latitude interpolation on the AIS data of the three-dimensional route density scatter image in the S1 in a three-dimensional density space to obtain ship dense route information on different elevation data, identifying target elevation data of the ship dense route information, carrying out inversion and putting the target elevation data back to a two-dimensional coordinate system, and obtaining a ship dense route two-dimensional plane graph in a continuous space; and S3, converting the coordinate system of the two-dimensional plane graph in the step S2 into a longitude and latitude coordinate system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a simple and efficient method for extracting ship routes, in particular to an intelligent method for extracting ship routes oriented to track images. Background Art

[0002] Ship route extraction can provide effective route planning and navigation safety, so it has important application value in the field of navigation. At present, the research on ship route extraction mainly focuses on traditional route design based on ocean data and rules. This method is often time-consuming and labor-intensive, and is subject to some limitations when designing large-scale routes such as long-distance routes. How to use the historical tracks of ships to carry out intelligent route design in batches is a new idea for current route extraction methods.

[0003] In the field of ship route extraction, there are currently the following main schools of thought. The first is a method based on empirical rules. This method relies on the knowledge and experience of the ship driver, takes into account factors such as ocean characteristics, weather, waterways and ship performance, and uses subjective experience to determine the ship route. The second is a method based on oceanography. This method uses traditional oceanographic knowledge and technology to analyze ocean currents, meteorology, topography and other factors, and combines historical navigation data for route planning. The third is a mathematical model method based on AIS data. Through the original AIS big data, it simulates ship navigation and the ocean environment to determine the route, and uses an optimization algorithm to solve it. In addition, there is a fourth method that performs ship trajectory clustering extraction based on two-dimensional ship route images to achieve information extraction of ship routes.

[0004] Several existing schools of thought have the following shortcomings when facing complex marine environments and large amounts of data: methods based on empirical rules are too dependent on the experience of sailors, are easily limited by subjective judgment and individual differences, and cannot fully utilize information such as big data and ocean forecasts; although methods based on traditional oceanography take into account marine environmental factors, they have certain limitations when dealing with variable and nonlinear marine data; methods based on mathematical models not only require the establishment of complex mathematical models, but also have high requirements for the selection and calibration of model parameters. In addition, AIS big data is expensive to purchase and difficult to obtain; the problem with the route extraction method based on two-dimensional image processing is that it is impossible to accurately determine the boundaries of different routes, and errors are difficult to eliminate. In the image processing mode, the first problem with machine learning and deep learning methods is that the processing process is complex and the calculation is difficult; secondly, the problem with two-dimensional grayscale image processing methods is that the semantic threshold segmentation is unclear, there is a lot of noise, and it is difficult to improve the processing effect, and with each iterative processing, the error will increase. Summary of the invention

[0005] The present invention solves the following problems existing in various ship route extraction methods at present: methods relying on the experience of navigators are easily limited by subjective judgment and individual differences, and cannot make full use of information such as big data and ocean forecasts; methods based on traditional oceanography have limitations in processing variable and nonlinear ocean data; methods based on mathematical models have complex mathematical models, high requirements for model parameter selection and calibration, AIS big data is expensive to purchase and difficult to obtain; route extraction methods based on two-dimensional image processing have many noises, cannot accurately determine the boundaries of different routes, and increase the error due to iteration; machine learning and deep learning methods have complex processing processes and high computational difficulty, and it is difficult to improve the processing effect. A ship route extraction method oriented to track images is provided, which can automatically extract a large amount of ship route information without being affected by subjective factors, can completely overcome the influence of noise, can perform high-precision identification and extraction of ship routes of different busyness, does not require complex mathematical models and requires high parameter selection, saves human resources, improves work efficiency, and makes the extraction and analysis process of route information more efficient.

[0006] The technical solution of the present invention is:

[0007] The ship route extraction method for trajectory images includes the following steps:

[0008] S1: For the route heat map image formed by the ship AIS data, a three-dimensional data extraction method is used to convert it into a three-dimensional route density scatter image;

[0009] S2: interpolate the AIS data of the three-dimensional route density scatter image of S1 in latitude in the three-dimensional density space to obtain the dense ship route information at different elevation data, identify its target elevation data and invert it back to the two-dimensional coordinate system to obtain a two-dimensional plane diagram of the dense ship route in continuous space;

[0010] S3: Convert the coordinate system of the two-dimensional plane map in step S2 into a longitude and latitude coordinate system.

[0011] S1 The three-dimensional data extraction method comprises the following steps:

[0012] S11: Obtain the ship AIS route heat map;

[0013] S12: By calculating equations (1)-(3), a three-dimensional continuous scattered point density ship route map, i.e., a three-dimensional route density scattered point map, is obtained:

[0014] X=x (14)

[0015] Y=y (15)

[0016] Z=I_r (16)

[0017] Where:

[0018] x is the horizontal coordinate of the plane graph; y is the vertical coordinate of the plane graph; X is the horizontal coordinate of the three-dimensional scatter plot; Y is the vertical coordinate of the three-dimensional scatter plot; I_r is the red component of the plane graph.

[0019] S2 interpolation includes the following steps:

[0020] The bilinear interpolation algorithm is used to perform linear interpolation on the two independent directions X and Y that are perpendicular to each other in step S12, respectively. Weights are assigned according to the distance from the interpolation point, and the values ​​of the four nearest neighboring points are weighted averaged to obtain the elevation estimate of the interpolation point and obtain the Z-axis elevation data image. The interpolation expression in the X direction is:

[0021]

[0022] The interpolation expression in the Y direction is:

[0023]

[0024] In formulas (4)-(6), Q 11 , Q 21 , Q 12 , Q 22 is the pixel point, x, x1, x2 are the coordinate boundaries, f(x, y1) and f(x, y2) are the pixel values ​​of R1 and R2;

[0025] S2 also uses edge detection and edge smoothing technology to process the image boundaries of the two-dimensional plane map, replacing the width of the ship's navigation track with a color boundary, thereby obtaining the location information of the customary route boundary.

[0026] S3 is converted to the longitude and latitude coordinate system according to the latitude growth rate formula, that is, based on the Mercator projection method, the pixel coordinates are converted into the longitude and latitude coordinate system, and then the latitude growth rate algorithm is used to quickly extract information for navigation use. Specifically:

[0027] Assuming that the origin of the Mercator projection coordinate system is (0, λ0), the X-axis is the equator, and the Y-axis is perpendicular to the equator at longitude λ0, the Mercator projection formula is:

[0028] x=λ-λ0 (20)

[0029]

[0030] λ=x+λ0 (23)

[0031]

[0032] Where λ is the longitude, is the latitude. Formulas (7) and (8) are positive calculations, and formulas (9) and (10)

[0033] For the inverse operation, the formula for the latitude growth rate is (11).

[0034] The ship route extraction method for trajectory images further includes step S4:

[0035] S4: Route reconstruction: Link the different three-dimensional route density scatter images obtained from different AIS data or the track points of the long-distance data source of the ship route, that is, associate the route information to form a longer route segment.

[0036] The association of the route reconstruction of S4 is to determine whether they are associated by setting a distance threshold between pixels. If they are associated, they are connected, and if they are not associated, they are removed. The formula is as follows (12):

[0037]

[0038] Among them, x A and A is the coordinate of the previous pixel, x B and B is the coordinate of the latter pixel, and then decides whether to associate the two pixels by applying a threshold to the distance.

[0039] The route reconstruction in step S4 also includes route smoothing processing. Formula (13) is the route smoothing processing formula, which is as follows:

[0040]

[0041] Among them, G(x,y) is the output pixel value of the Gaussian filter, x and y are the offsets of the pixel relative to the center pixel, is the standard deviation of the Gaussian distribution, which controls the smoothness of the filter.

[0042] A storage medium, characterized in that it stores computer instructions, which are used to implement the ship route extraction method for trajectory images as described in any one of claims 1 to 8 when executed by a processor.

[0043] Technical effects:

[0044] The present invention converts the AIS ship route heat map into a three-dimensional route density scatter image through a three-dimensional data extraction method, that is, the two-dimensional trajectory image is upgraded to a three-dimensional image, and then the ship dense route information on different elevation data is obtained from the three-dimensional route density scatter map, and the elevation data of the target route is identified, thereby effectively enhancing the route visualization ability and discrimination and achieving zero-error information extraction, which can more accurately represent the navigation trajectory of the ship in the ocean, and then by re-converting the trajectory image into two dimensions and using image processing technology to convert the two-dimensional plane map coordinate system into a latitude and longitude coordinate system, the route information can be extracted.

[0045] The present invention is not affected by the subjective judgment and individual differences of sailors, can make full use of AIS data, and extract a large amount of ship route information. At the same time, the method of the present invention does not involve complex mathematical models, does not need to select and calibrate model parameters, overcomes the influence of noise, can accurately determine the boundaries of different routes, does not need iteration, and therefore does not increase errors. The processing process is simple, the calculation difficulty is low, and the processing effect is good. The present invention not only saves human resources and improves work efficiency, but also makes the extraction and analysis process of route information more efficient. The method of the present invention can overcome the shortcomings of various existing ship route extractions, has advantages in processing large-scale ship trajectory data, and can provide more accurate and efficient ship route information.

[0046] The three-dimensional data extraction method of the present invention is simple, and a ship route map with three-dimensional continuous scattered point density can be obtained through simple calculation.

[0047] Through the bilinear interpolation fitting algorithm, we successfully obtained the Z-axis elevation data images of different elevations (colors), obtained the ship's usual route information, and thus obtained the relevant data of the target route. By using visual mathematical tools, the route data of different elevation positions in the entire space can be presented simultaneously; that is, bilinear interpolation can fill the blank areas between pixels in the image, assign weights according to the distance from the interpolation point, and perform weighted averaging of the values ​​of the four nearest neighboring points to obtain the elevation estimate of the interpolation point, realize further perception of the accurate location information of the route, and perform high-precision identification and extraction of ship routes of different busy levels. That is, the bilinear interpolation algorithm utilizes multi-dimensional information, which helps to improve the accuracy of the interpolation results.

[0048] The method of the present invention performs image processing on the basis of the ship navigation track data image, and then inverts the ship route data in a two-dimensional plane map. The method mainly replaces the ship navigation track width with a color boundary, and identifies the boundaries of routes with different busyness or different ship route densities in the image color transition area, so as to accurately determine the boundary information of the set route.

[0049] Next, the route reconstruction pixels of the present invention are judged whether they are associated by setting a distance threshold between pixel points. If they are associated, they are connected, and if they are not associated, they are eliminated. In this way, in areas with higher density, the algorithm can generate denser routes, and in areas with lower density, it can generate sparser routes to better reflect the distribution of route density. Therefore, the route reconstruction technology is based on the changes in the ship density heat map, identifies the route width and extracts the route direction, adaptively adjusts the route precision, and then determines the strongly associated waypoints to achieve track path expression.

[0050] Finally, the route reconstruction algorithm introduces a smoothing technique to make the generated route path smoother and more continuous. The present invention uses a Gaussian filter method to achieve this in order to reduce noise and unnecessary jitter.

[0051] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flowchart of the present invention;

[0053] Figure 2 It is a schematic diagram of bilinear interpolation of the present invention;

[0054] Figure 3 It is the technical roadmap of the present invention, i.e. the full flow chart;

[0055] Figure 4 This is the Qiongzhou Strait Experiment Application Map;

[0056] Figure 5 It is an experimental application diagram of the Arctic route;

[0057] Figure 6 It is a latitude and longitude information table obtained by extracting information of the Qiongzhou Strait through the present invention. DETAILED DESCRIPTION

[0058] Figure 1 In the paper, the ship route extraction method for trajectory images mainly solves the problem of route design and route planning before the ship sails. It focuses on key technologies such as image information stratification, image information extraction, route information conversion, and route information recognition and storage. It includes the following steps:

[0059] S1: For the route heat map image formed by the ship AIS data, a three-dimensional data extraction method is used to convert it into a three-dimensional route density scatter image, that is, heat map acquisition and coordinate dimension upgrading: the AIS ship route heat map data is converted into a three-dimensional ship density scatter map through a three-dimensional extraction method: the AIS ship route heat map is converted based on pixel color data in the three-dimensional dimension of elevation space, and the route data under each density dimension is obtained through the image color recognition algorithm: as shown in formula 1-3, combined with the collected two-dimensional ship route density heat map, a three-dimensional route density scatter map with elevation information is obtained.

[0060] X=x (27)

[0061] Y=y (28)

[0062] Z=I_r (29)

[0063] Where:

[0064] x is the horizontal coordinate of the plane graph; y is the vertical coordinate of the plane graph; X is the horizontal coordinate of the three-dimensional scatter graph; Y is the vertical coordinate of the three-dimensional scatter graph, Z is the vertical coordinate of the three-dimensional scatter graph; I_r is the red component of the plane graph.

[0065] S2: Interpolate the AIS data of the three-dimensional route density scatter image obtained in step S1 in latitude in the three-dimensional density space to obtain the dense ship route information at different elevation data, identify its target elevation data and invert it back to the two-dimensional coordinate system to obtain a two-dimensional plane diagram of the dense ship route in continuous space: In the face of large errors in single-dimensional linear interpolation, a bilinear interpolation algorithm is used to perform linear interpolation in two independent directions. The bilinear interpolation algorithm utilizes multi-dimensional information, which helps to improve the accuracy of the interpolation result (see Figure 2 ).

[0066] Linear interpolation is performed on the two independent directions X and Y perpendicular to each other in step S1, weights are assigned according to the distance from the interpolation point, and the values ​​of the four nearest neighboring points are weighted averaged to obtain the elevation estimate of the interpolation point, that is, the Z-axis elevation selection to obtain the Z-axis elevation data image;

[0067] The interpolation expression in the X direction is:

[0068]

[0069] The interpolation expression in the Y direction is:

[0070]

[0071] In formulas (4)-(6), Q 11 , Q 21 , Q 12 , Q22 is the pixel point, x, x1, x2 are the coordinate boundaries, f(x, y1) and f(x, y2) are the pixel values ​​of R1 and R2.

[0072] Identify the target elevation data, that is, extract and collect the target elevation information of the habitual route and invert it back to the two-dimensional coordinate system to obtain a two-dimensional plane map of dense ship routes in continuous space; further, perform image processing on the two-dimensional plane image according to the calculated route, including grayscale processing, Gaussian filter smoothing processing, and use edge detection and edge smoothing technology to process the image boundary of the two-dimensional plane map, and replace the width of the ship's navigation track with a color boundary to obtain the location information of the habitual route boundary. Edge detection uses improved Sobel edge detection, edge smoothing processing includes edge square sum and square root processing, and noise processing is also performed.

[0073] Replacing the width of a ship's navigation track with a color boundary refers to inverting the ship's route data to form a two-dimensional plane map, and identifying the boundaries of routes with different busy levels or different ship route densities based on the color transition area of ​​the image, so as to accurately determine the boundary information of the set route.

[0074] S3: Convert the coordinate system of the two-dimensional plane map in step S2 into a latitude and longitude coordinate system:

[0075] Under the high-density image information coordinates, the coordinate system conversion under the navigation chart is carried out to realize the extraction of route data. Based on the two-dimensional coordinate system of the multi-color image (that is, the two-dimensional plane map formed by inversion), it is converted into the longitude and latitude coordinate system according to the latitude gradual growth rate formula.

[0076] Based on the Mercator projection method, pixel coordinates are converted into longitude and latitude coordinate systems for navigation. Assume that the origin of the Mercator projection coordinate system is (0, λ0), the X axis is the equator, and the Y axis is perpendicular to the equator at longitude λ0. The Mercator projection formula is:

[0077] x=λ-λ0 (33)

[0078]

[0079]

[0080] λ=x+λ0 (36)

[0081]

[0082] Where λ is the longitude, is the latitude. Formula (7) and (8) are positive calculations, formula (9) and (10) are inverse calculations, and the formula for the latitude growth rate is formula (11).

[0083] After extracting information from the dense routes of AIS ships based on the bilinear interpolation algorithm, the two-dimensional plane coordinate system is converted into a longitude and latitude coordinate system in combination with the Mercator coordinate system principle, and the longitude and latitude of the dense target routes are converted. Then error identification is performed, and information is automatically extracted through route reconstruction. The technical route of the whole process can be found in Figure 3 .

[0084] S4 Route reconstruction: Route reconstruction technology links the trajectory points of different three-dimensional route density scatter images or long-distance data sources of ship routes obtained from different AIS data, that is, to associate route information to form a longer route segment. In order to improve the visualization and analysis effect, the route reconstruction algorithm can introduce smoothing technology to make the generated route path smoother and continuous. The present invention adopts the Gaussian filter method to achieve this, so as to reduce noise and unnecessary jitter. The route reconstruction technology of the present invention adaptively adjusts the refinement of the route according to the changes in the different ship density heat maps obtained from different AIS data. In areas with higher density, the algorithm can generate denser routes, and in areas with lower density, it can generate sparser routes to better reflect the distribution of route density and realize automatic extraction. The distance threshold is set between pixels to determine whether they are associated. If there is an association, it is connected, and if there is no association, it is eliminated. The formula is as shown in (12). Formula (13) is the route smoothing processing formula. Formula (14) is the route smoothing processing formula.

[0085]

[0086] Among them, x A and A is the coordinate of the previous pixel, x B and B is the coordinate of the latter pixel, and then decides whether to associate two pixels by applying a threshold to the distance between the pixels.

[0087]

[0088] Where G(x,y) is the output pixel value of the Gaussian filter. x and y are the offsets of the pixel relative to the center pixel. is the standard deviation of the Gaussian distribution, which controls the smoothness of the filter.

[0089] The present invention also includes a storage medium, characterized in that it stores computer instructions, and when the computer instructions are executed by a processor, they are used to implement the ship route extraction method for trajectory images as described above.

[0090] Specific application cases:

[0091] Taking the heat maps of the Qiongzhou Strait and Arctic routes of the total AIS data for the whole year of 2022 as an example, information on the customary routes of ships in narrow waterways and sea areas was extracted respectively.

[0092] First, the AIS ship route map is obtained. Through the simple calculation of formulas (1)-(3), a three-dimensional continuous scattered point density ship route map, that is, a three-dimensional route density scattered point image, is obtained. Bidirectional linear interpolation is performed in the X and Y directions to obtain the Z-axis elevation data image. The Z value of the target route is selected by elevation to obtain the ship route data in the continuous space of the Arctic. The thermal image presented by the AIS data collected in the Qiongzhou Strait at latitudes of 19.9798°-20.2866° and longitudes of 109.8344°-110.3934° is three-dimensionally modeled according to the simple calculation of formulas (1)-(3) to obtain elevation information, such as Figure 4 (a)-(b) show; extract and collect the target elevation information (data) of the customary route, such as Figure 4 (c) shows that the two-dimensional information is inverted and returned to the two-dimensional image, as shown in Figure 4 (d) as shown.

[0093] Secondly, the two-dimensional image is processed according to the calculated route, including edge detection and edge smoothing, and the width of the ship's navigation track is replaced by a color boundary to obtain the boundary of the customary route, such as Figure 4 (e) as shown.

[0094] Finally, after the pixel coordinates under the influence of latitude growth rate are converted to the Mercator projection coordinate system to obtain information that can be used for navigation, the information is extracted through the fast latitude growth rate formula automatic extraction information technology, such as Figure 6 The latitude and longitude information is shown in Table 1. The processed information is compared with the original information, and the error is within 5%. Figure 4 As shown in (f)-(g), Figure 4 (j) is the ship route boundary.

[0095] Similarly, experiments were conducted in Arctic offshore waters. Figure 5 shown. Figure 5 (g) is a customary route, Figure 5 (h) For this busy route, there are many rest routes due to the sparse Arctic routes.

Claims

1. Ship route extraction method for trajectory images, Its characteristics are The following steps are involved: S1: For the route heat map image formed by the ship AIS data, a three-dimensional data extraction method is used to convert it into a three-dimensional route density scatter image; S2: interpolate the AIS data of the three-dimensional route density scatter image of S1 in latitude in the three-dimensional density space to obtain the dense ship route information at different elevation data, identify its target elevation data and invert it back to the two-dimensional coordinate system to obtain a two-dimensional plane diagram of the dense ship route in continuous space; S3: Convert the coordinate system of the two-dimensional plane map in step S2 into a longitude and latitude coordinate system.

2. The ship route extraction method for trajectory images according to claim 1, Its characteristics are S1 The three-dimensional data extraction method comprises the following steps: S11: Obtain the ship AIS route heat map; S12: By calculating equations (1)-(3), a three-dimensional continuous scattered point density ship route map, i.e., a three-dimensional route density scattered point map, is obtained: X=x (1) Y=y (2) Z=I_r (3) Where: x is the horizontal coordinate of the plane graph; y is the vertical coordinate of the plane graph; X is the horizontal coordinate of the three-dimensional scatter plot; Y is the vertical coordinate of the three-dimensional scatter plot; I_r is the red component of the plane graph.

3. The ship route extraction method for trajectory images according to claim 2, Its characteristic is S2 Interpolation involves the following steps: The bilinear interpolation algorithm is used to perform linear interpolation on the two independent directions X and Y that are perpendicular to each other in step S12, respectively. Weights are assigned according to the distance from the interpolation point, and the values ​​of the four nearest neighboring points are weighted averaged to obtain the elevation estimate of the interpolation point and obtain the Z-axis elevation data image. The interpolation expression in the X direction is: The interpolation expression in the Y direction is: In formulas (4)-(6), Q 11 , Q 21 , Q 12 , Q 22 is a pixel, x, x 1 、x 2 is the coordinate boundary, f(x,y 1 ) and f(x,y 2 ) is R 1 , R 2 The pixel value of .

4. The ship route extraction method for trajectory images according to claim 3, Its characteristic is S2 Edge detection and edge smoothing techniques are also used to process the image boundaries of the two-dimensional plane map, replacing the width of the ship's navigation track with a color boundary to obtain the location information of the customary route boundary.

5. The ship route extraction method for trajectory images according to claim 4, Its characteristics are S3 is converted to the longitude and latitude coordinate system according to the latitude growth rate formula, that is, based on the Mercator projection method, the pixel coordinates are converted into the longitude and latitude coordinate system, and then the latitude growth rate algorithm is used to quickly extract information for navigation use. Specifically: Assume that the origin of the Mercator projection coordinate system is (0, λ 0 ), the X-axis is the equator, and the Y-axis is at longitude λ 0 The Mercator projection formula is perpendicular to the equator: x=λ-λ 0 (7) λ=x+λ 0 (10) Where λ is the longitude, is the latitude. Formulas (7) and (8) are positive calculations, and formulas (9) and (10) For the inverse operation, the formula for the latitude growth rate is (11).

6. The ship route extraction method for trajectory images according to claim 5, Its characteristics are The step S4 is also included: S4: Route reconstruction: Link the different three-dimensional route density scatter images obtained from different AIS data or the track points of the long-distance data source of the ship route, that is, associate the route information to form a longer route segment.

7. The ship route extraction method for trajectory images according to claim 6, Its characteristics are The association of the route reconstruction of S4 is to determine whether they are associated by setting a distance threshold between pixels. If they are associated, they are connected, and if they are not associated, they are removed. The formula is as follows (12): Among them, x A and A is the coordinate of the previous pixel, x B and B is the coordinate of the latter pixel, and then a threshold is applied to the distance d to decide whether to associate the two pixels.

8. The ship route extraction method for trajectory images according to claim 7, Its characteristics are The route reconstruction in step S4 also includes route smoothing processing, the formula is as follows: Among them, G(x,y) is the output pixel value of the Gaussian filter, x and y are the offsets of the pixel relative to the center pixel, is the standard deviation of the Gaussian distribution, which controls the smoothness of the filter.

9. A storage medium, Its characteristic is storage There are computer instructions, which, when executed by a processor, are used to implement the ship route extraction method for trajectory images as described in any one of claims 1-8.

Citation Information

Cited By

  • Ship route optimization method and system based on environmental space-time change

    CN122590867A