A Method and System for Calculating Multi-Dimensional Features of Rivers in Remote Sensing Images

Abstract

The present invention provides a method and system for calculating multi-dimensional features of rivers in remote sensing images. First, the river area and the river boundary line are extracted from the target image. Based on the river area, the river center line is obtained, and a plurality of center points are evenly generated along the river center line. Then, for each center point, the corresponding river cross-section line is determined, and a plurality of cross-section points are evenly generated along each river cross-section line. Finally, a preset interpolation method is used to interpolate the feature values of the cross-section points into the river area to generate river raster data, and each pixel of the river raster data has at least the river width feature, the river channel direction feature, and the off-shore distance feature. The present invention can quickly and accurately calculate the river width, the off-shore distance of water pixels, and the river channel angle, overcoming the deficiencies of traditional methods in terms of accuracy and efficiency. It is applicable to the processing of large-scale river images, meets the needs of large-scale data analysis, and can calculate multiple features of rivers simultaneously, providing more comprehensive data support.

Description

Technical Field

[0001] The present invention belongs to the field of remote sensing, and particularly relates to a method and system for calculating multi-dimensional features of rivers in remote sensing images. Background Art

[0002] With the rapid development of remote sensing technology, remote sensing images are increasingly widely used in the fields of water resource management, environmental monitoring, disaster prediction, etc. As an important water body form, the calculation of river width and river channel direction is of great significance for hydrological analysis, ecological protection, and urban planning. The water body's distance from the shore helps to evaluate the interaction between the river and the surrounding environment, especially providing important data support in aspects such as water flow dynamics, sediment transport, and water quality changes. Precise calculation of the water body's distance from the shore can provide strong guarantees for water resource management, flood warning, and wetland protection. Due to the wide coverage and rich historical data of remote sensing images, as well as the advantage of efficient data acquisition, it has become an important tool for studying river characteristics.

[0003] The calculation of river width, water body's distance from the shore, and river channel direction usually relies on ground measurements or traditional image processing techniques. Although ground measurements have high precision, they are cumbersome, time-consuming, and limited by space and time, making it difficult to meet the measurement requirements for large areas or high frequencies. Traditional remote sensing image river width calculation methods, such as RivWidth, RivWidthCloud, RivaMap, and RiverWidthIE, mainly rely on digital image processing techniques or combined with neural networks to extract the river channel centerline and calculate the width through multiple convolution operations. These methods are effective under certain conditions, but in cases where the terrain is complex or the river channel morphology changes significantly, the recognition accuracy may decrease. At the same time, the large amount of calculation leads to low processing efficiency, making it difficult to meet the needs of large-scale, high-resolution remote sensing data. In addition, existing remote sensing image analysis mostly focuses on single-index extraction, lacking comprehensive analysis of multi-dimensional features of rivers (such as width, water body pixel's distance from the shore, river channel direction). Due to the complex and variable hydrological, geographical, and climatic characteristics of different rivers, the applicability of existing technologies in multi-dimensional calculations still has certain limitations.

[0004] Therefore, improving the accuracy and efficiency of calculating river characteristics in remote sensing images, especially in multi-dimensional feature calculations, is still an urgent problem to be solved in the field of remote sensing image analysis. Summary of the Invention

[0005] The present invention provides a method and system for calculating multi-dimensional features of rivers in remote sensing images to solve the problems in the prior art that the accuracy and efficiency of calculating river characteristics are relatively low, and it focuses on single-index extraction, lacking comprehensive analysis of multi-dimensional features of rivers, such as width, water body pixel's distance from the shore, and river channel direction.

[0006] To solve the above technical problems, the embodiments of the present invention disclose the following technical solutions:

[0007] One aspect of the present invention provides a method for calculating multi-dimensional features of rivers in remote sensing images, including:

[0008] Extract the river area and the river boundary line in the target image;

[0009] Obtain the river center line based on the river area, and uniformly generate a plurality of center points along the river center line;

[0010] For each center point, determine the corresponding river cross-section line;

[0011] Uniformly generate a plurality of cross-section points along each river cross-section line. The cross-section points have at least river width features, river channel direction features, and off-shore distance features. Among them, the length of the river cross-section line corresponding to the cross-section point is used as the river width feature value, the river channel angle of the cross-section point is calculated as the river channel direction feature value, and the distance between the cross-section point and the river boundary line is calculated as the off-shore distance feature value;

[0012] Use a preset interpolation method to interpolate the feature values of the cross-section points into the river area to generate river raster data. Each pixel of the river raster data has at least river width features, river channel direction features, and off-shore distance features.

[0013] Optionally, the extracting the river area and the river boundary line in the target image includes:

[0014] Identify the water body area in the target image as the river area, and convert the river area into polygon vector data;

[0015] Preprocess the polygon vector data and extract the river boundary line.

[0016] Optionally, the preprocessing the polygon vector data and extracting the river boundary line includes:

[0017] Use the keep key bends geometric simplification algorithm to simplify the polygon vector data;

[0018] Identify and fill the void areas or fracture areas in the polygon vector data;

[0019] Use the polynomial approximation method based on the exponential kernel to smooth the polygon vector data;

[0020] Extract the river boundary line based on the polygon vector data after the above preprocessing is completed.

[0021] Optionally, the method further includes:

[0022] Generate a plurality of boundary points uniformly along the river boundary line with a first preset step size.

[0023] Optionally, before performing the step of determining the corresponding river cross-section line for each center point, the method further includes:

[0024] Obtain the minimum distance d_min between the center point and all boundary points;

[0025] Connect two adjacent center points to the center point, and determine the river cross-section direction corresponding to the center point according to the orthogonal direction of the connection line;

[0026] Calculate the azimuth angle between the river cross-section direction and the preset direction .

[0027] Optionally, the step of determining the corresponding river cross-section line for each center point includes:

[0028] Taking the center point as the starting point, extend lines in both directions along the corresponding river cross-section direction, and the length of the extension line is a preset multiple of d_min;

[0029] Retain the line segment of the extension line within the river area as the river cross-section line corresponding to the center point.

[0030] Optionally, the calculation of the river channel angle of the cross-section point as the river channel direction eigenvalue includes:

[0031] Calculate the river channel angle of the cross-section point using the following formula :

[0032]

[0033]

[0034] where saa is the solar azimuth angle at the cross-section point, obtained from the metadata of the target image; is the azimuth angle of the river cross-section direction corresponding to the cross-section point.

[0035] Optionally, the calculation of the distance between the cross-section point and the river boundary line as the offshore distance eigenvalue includes:

[0036] Obtain two intersection points of the river cross-section line corresponding to the cross-section point and the river boundary line;

[0037] Calculate the distances between the cross-section point and the two intersection points respectively as the offshore distance eigenvalue.

[0038] Optionally, the step of interpolating the eigenvalue of the cross-section point into the river area using a preset interpolation method includes:

[0039] Interpolate the eigenvalue of each cross-section point into the river area using the inverse distance weighted interpolation method.

[0040] Another aspect of the present invention provides a remote sensing image river multi-dimensional feature calculation system, including:

[0041] An extraction module configured to extract a river area and a river boundary line in a target image;

[0042] A center point module configured to obtain a river center line based on the river area and uniformly generate a plurality of center points along the river center line;

[0043] A river channel cross-section line module configured to determine a corresponding river channel cross-section line for each center point;

[0044] A cross-section point module configured to uniformly generate a plurality of cross-section points along each river cross-section line, where the cross-section points have at least a river width feature, a river channel direction feature, and an offshore distance feature. Among them, the length of the river cross-section line corresponding to the cross-section point is used as the river width feature value, the river channel angle of the cross-section point is calculated as the river channel direction feature value, and the distance between the cross-section point and the river channel boundary line is calculated as the offshore distance feature value;

[0045] An interpolation module configured to interpolate the eigenvalue of the cross-section point into the river area using a preset interpolation method to generate river raster data, and each pixel of the river raster data has at least a river width feature, a river channel direction feature, and an offshore distance feature.

[0046] A method and system for calculating multi-dimensional features of a river in a remote sensing image disclosed by the present invention. First, extract a river area and a river boundary line in a target image, obtain a river center line based on the river area, and uniformly generate a plurality of center points along the river center line; then, for each center point, determine a corresponding river channel cross-section line; uniformly generate a plurality of cross-section points along each river cross-section line, and the cross-section points have at least a river width feature, a river channel direction feature, and an offshore distance feature; finally, interpolate the eigenvalue of the cross-section point into the river area using a preset interpolation method to generate river raster data, and each pixel of the river raster data has at least a river width feature, a river channel direction feature, and an offshore distance feature.

[0047] A method and system for calculating multi-dimensional features of a river in a remote sensing image disclosed by the present invention can quickly and accurately calculate the river width, the offshore distance of water body pixels, and the river channel angle, overcoming the deficiencies of traditional methods in terms of accuracy and efficiency. It is applicable to large-scale river image processing, meets the needs of large-scale data analysis, and can calculate multiple features of a river simultaneously, providing more comprehensive data support. In addition, the method and system provided by the present invention can also reduce manual intervention and improve the data processing efficiency and reliability.

[0048] The Summary of the Invention section is provided to introduce, in a simplified form, the selection of concepts that will be further described in the Detailed Description below. The Summary of the Invention section is not intended to identify the key features or essential features of the present disclosure, nor is it intended to limit the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The above and other objects, features, and advantages of the present disclosure will become more apparent by describing the exemplary embodiments of the present disclosure in more detail in conjunction with the accompanying drawings, wherein, in the exemplary embodiments of the present disclosure, the same reference numerals generally represent the same components.

[0050] Figure 1 It is a schematic flowchart of a method for calculating multi-dimensional features of rivers in remote sensing images provided by an embodiment of the present invention;

[0051] Figure 2 It is a schematic flowchart of implementing Figure 1 step S100 in the present invention;

[0052] Figure 3 It is a schematic flowchart of implementing Figure 2 step S102 in the present invention;

[0053] Figure 4 It is a schematic flowchart of another method for calculating multi-dimensional features of rivers in remote sensing images provided by an embodiment of the present invention;

[0054] Figure 5 It is a schematic flowchart of implementing Figure 1 step S300 in the present invention;

[0055] Figure 6 It is a schematic structural diagram of a system for calculating multi-dimensional features of rivers in remote sensing images provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0057] 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 "at least partially based 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. There may also be other explicit and implicit definitions hereinafter.

[0058] Figure 1 The flowchart of a method for calculating multi-dimensional features of rivers in remote sensing images provided by an embodiment of the present invention is shown as Figure 1 shown, and the method includes the following steps:

[0059] Step S100: Extract the river area and the river boundary line in the target image.

[0060] Take a remote sensing image of the target river as the target image. In an embodiment disclosed in the present invention, as Figure 2 shown, the following sub-steps can be used to complete step S100:

[0061] Step S101: Identify the water body area in the target image as the river area and convert the river area into planar vector data.

[0062] In an embodiment disclosed in the present invention, an input mask file can be used to confirm the water body area, or the normalized difference water index (NDWI) can be used for water body identification. NDWI is calculated based on the reflectance difference between the green and near-infrared bands of the remote sensing image and can effectively distinguish water bodies from other ground objects. Specifically, the calculation formula of NDWI is:

[0063]

[0064] where Green represents the green band and NIR represents the near-infrared band. After calculation according to this formula, areas with higher NDWI values usually represent water bodies because water bodies have higher reflectance in the green band and lower reflectance in the near-infrared band. By calculating and setting an appropriate threshold, the water body area can be effectively identified. For example, areas with NDWI values greater than 0 can be considered as water bodies, i.e., river areas.

[0065] For example, if Sentinel-2 L1C level data is obtained, NDWI is calculated to identify water bodies:

[0066]

[0067] The part where NDWI > 0 is considered as the water body.

[0068] If Sentinel-2 Level 2A data is obtained, the water body area can be directly obtained through the scene classification band.

[0069] Converting the river area into polygon vector data using GIS software means converting pixel-based raster data into vector data with geometric shapes (such as polygons), thus obtaining vector data with precise geometric shapes and positions, laying the foundation for subsequent processing steps.

[0070] Step S102: Preprocess the polygon vector data and extract the river boundary line.

[0071] After converting the remote sensing image into vector data, the river polygon vector data usually has noise, irregular parts or missing areas, so further processing is required.

[0072] In an embodiment disclosed by the present invention, as Figure 3 shown, the following method can be used to complete step S102:

[0073] Step S1021: Simplify the polygon vector data using the key bend retention geometric simplification algorithm.

[0074] After vectorizing the river area, there will be redundant points and unnecessary details, resulting in a large amount of data and overly complex boundaries, which affect visualization and subsequent analysis. Therefore, in the embodiment of the present invention, through the key bend retention (Wang-Müller) geometric simplification algorithm, redundant details are removed, unnecessary small changes or small curves are reduced, while maintaining the river morphology, making the river boundary more concise and facilitating subsequent processing.

[0075] Step S1022: Identify and fill the void areas or fracture areas in the polygon vector data.

[0076] In the river polygon vector, voids or fractures may be caused by factors such as ships, bridges, floating objects, etc. These voids usually do not represent the real river boundary, so they need to be filled. To ensure the integrity of the river boundary, based on the topological structure of the vector data, the voids or gaps between polygon boundaries are identified, and the missing parts are automatically filled according to the adjacent boundaries and the shapes of the existing polygons. The filled river boundary becomes more continuous, avoiding the missing impact caused by obstacles or ships, etc.

[0077] Step S1023: Smooth the polygon vector data using the polynomial approximation method based on the exponential kernel.

[0078] The boundaries of rivers are sometimes affected by artificial structures (such as docks, bridges, etc.), resulting in uneven or deformed boundaries. By using the Polynomial Approximation based on Exponential Kernel (PAEK) method, the sharp angles and unnatural shapes caused by image noise or artificial structures are removed, making the river boundaries more conform to the shape of natural water bodies.

[0079] Step S1024: Extract the river boundary line based on the planar vector data after the above preprocessing.

[0080] After completing the simplification, gap filling, and smoothing processes, use GIS software or other contour extraction tools to extract the accurate river boundary line vector from the processed river surface vector. The river boundary line after the above processing is more accurate and coherent, and can provide accurate data support for subsequent steps.

[0081] In an embodiment disclosed by the present invention, after extracting the river boundary line, a plurality of boundary points need to be evenly generated along the river boundary line at a first preset step size.

[0082] For example, along the river boundary line vector generated in the previous step, a series of boundary points are evenly generated with a first preset step size of 10 meters. These points will be distributed along the river boundary and provide basic data for subsequent analysis.

[0083] Step S200: Obtain the river centerline based on the river area and evenly generate a plurality of center points along the river centerline.

[0084] Based on the river planar vector data generated in the previous step, extract and determine the centerline of the river. The centerline represents the main water flow path of the river and is located at the geometric center of the river.

[0085] A series of center points are evenly generated along the river centerline. For example, a center point is generated every 100 meters along the centerline.

[0086] In an embodiment disclosed by the present invention, before performing step S300 to determine the corresponding river cross-section line for each center point, as Figure 4 shown, the method disclosed by the present invention further includes the following steps:

[0087] Step S031: Obtain the minimum distance d_min between the center point and all boundary points.

[0088] For each center point p0, assume its coordinates in the image are (x0, y0), and the image coordinate system is the projection coordinate system. Calculate the minimum distance between this center point p0 and all river boundary points, that is, the shortest distance d_min from p0 to the river boundary.

[0089] Step S032: Connect two central points adjacent to the central point, and determine the river channel cross-section direction corresponding to the central point according to the orthogonal direction of the connecting line.

[0090] Find two central points p1(x1, y1) and p2(x2, y2) that are closest to p0, connect p1p2 to generate a straight line, determine the orthogonal direction of this straight line, and use it as the river channel cross-section direction corresponding to p0.

[0091] Step S033: Calculate the azimuth angle between the river channel cross-section direction and the preset direction .

[0092] In the embodiment disclosed in the present invention, calculate the angle between the river channel cross-section and the due north direction as the azimuth angle , and the calculation formula is as follows:

[0093]

[0094] It can describe the bending shape of the river, and the range is (0, 180) degrees.

[0095] Step S300: For each central point, determine the corresponding river channel cross-section line.

[0096] In an embodiment disclosed in the present invention, as Figure 5 shown, the following sub-steps can be used to complete step S300:

[0097] Step S301: Use the central point as the starting point, and make extension lines along the corresponding river channel cross-section direction to both sides respectively.

[0098] Based on the river central points generated in the previous steps, for each central point p0, extend it to both sides along the river channel cross-section direction, and the length of the extension line is a preset multiple of d_min. For example, the length of the extension line is d_min 1.5.

[0099] Step S302: Retain the line segments of the extension lines within the river area as the river channel cross-section lines corresponding to the central points.

[0100] During the extension process, only retain the partial line segments that intersect with the river planar vector, and eliminate the line segments with less than 2 intersection points with the river channel boundary line, so as to obtain effective river channel cross-section lines. Each cross-section line inherits all the features of its corresponding river central point. For each cross-section line, calculate its line segment length, and this length is the river width (width) corresponding to this cross-section line. In addition, define the projections of the two end points of the cross-section line in the x direction of the central point local coordinate system. The end point with a positive projection is the positive end point p_point, representing the positive bank of the river channel; the end point with a negative projection is the negative end point n_point, representing the negative bank of the river channel.

[0101] Step S400: Uniformly generate a plurality of cross-section points along each river cross-section line.

[0102] Along each river cross-section line generated in the foregoing steps, a plurality of cross-section points are uniformly generated at a second preset step length (for example, 10 meters). Each cross-section point will inherit all the characteristics of the river cross-section line where it is located, that is, the cross-section point has at least the characteristics of river width, river channel direction, and offshore distance.

[0103] For each cross-section point, the characteristics of river width, river channel direction, and offshore distance are assigned to it. Among them, the length of the river cross-section line corresponding to the cross-section point is used as the river width characteristic value, the river channel included angle of the cross-section point is calculated as the river channel direction characteristic value, and the distance between the cross-section point and the river channel boundary line is calculated as the offshore distance characteristic value.

[0104] 1. The length of the river cross-section line where the cross-section point is located is used as the river width characteristic value.

[0105] 2. Obtain the solar zenith angle saa at the cross-section point. Then, establish a local coordinate system with the cross-section point as the origin. Assume that the projection of the solar ray direction on the ground is the positive direction of the x-axis, and use the right-hand rule to determine the xoy coordinate system. Define the included angle between the river cross-section line where the cross-section point is located and the x-axis as the river channel included angle rca. The river channel included angle is of great significance for analyzing the bending situation and water flow direction of the river, and its range is (-90, 90) degrees.

[0106] The river channel included angle of the cross-section point is calculated using the following formula :

[0107]

[0108]

[0109] Among them, saa is the solar azimuth angle at the cross-section point, which is obtained from the metadata of the target image; for example, the solar azimuth angle is read from the metadata of the image, and then bilinear interpolation is used to interpolate to a resolution of 10 meters to obtain the solar azimuth angle corresponding to each pixel in the image;

[0110] is the azimuth angle of the river cross-section direction corresponding to the cross-section point.

[0111] 3. In an embodiment disclosed in the present invention, the following method can be used to calculate the distance between the cross-section point and the river channel boundary line as the offshore distance characteristic value:

[0112] (1) Obtain two intersection points of the river cross-section line corresponding to the cross-section point and the river channel boundary line.

[0113] Obtain the intersection point p_point of the river cross-section line where the cross-section point is located and the positive bank of the river boundary line, and obtain the intersection point n_point of the river cross-section line where the cross-section point is located and the negative bank of the river boundary line.

[0114] (2) Calculate the distances between the cross-section point and the two intersection points respectively as the off-shore distance eigenvalue.

[0115] Calculate the distance between the cross-section point and the corresponding positive bank intersection point p_point to obtain the off-shore distance d1; the distance from the cross-section point to the corresponding negative bank intersection point n_point is the off-shore distance d2.

[0116] Step S500: Interpolate the eigenvalue of the cross-section point into the river area by using a preset interpolation method to generate river raster data.

[0117] In an embodiment disclosed by the present invention, the inverse distance weighted interpolation method can be used to interpolate the eigenvalue of each cross-section point into the entire river area. This process converts discrete point data into continuous area data through mathematical methods to generate a raster data model covering the entire river. Finally, the generated raster data will contain the river attribute data corresponding to each pixel (i.e., raster cell), namely river width feature, channel direction feature, off-shore distance feature, etc. These data can describe information such as the geometric shape, width, and off-shore distance of the river.

[0118] Figure 6 It is a schematic structural diagram of a remote sensing image river multi-dimensional feature calculation system disclosed by an embodiment of the present invention. As Figure 6 shown, the system includes the following modules:

[0119] Extraction module 1, configured to extract the river area and the river boundary line in the target image;

[0120] Center point module 2, configured to obtain the river center line based on the river area and uniformly generate a plurality of center points along the river center line;

[0121] River cross-section line module 3, configured to determine the corresponding river cross-section line for each center point;

[0122] Cross-section point module 4, configured to uniformly generate a plurality of cross-section points along each river cross-section line. The cross-section points have at least river width feature, channel direction feature, and off-shore distance feature. Among them, the length of the river cross-section line corresponding to the cross-section point is used as the river width eigenvalue, the channel included angle of the cross-section point is calculated as the channel direction eigenvalue, and the distance between the cross-section point and the river boundary line is calculated as the off-shore distance eigenvalue;

[0123] The interpolation module 5 is configured to interpolate the eigenvalue of the cross-section points into the river area by using a preset interpolation method to generate river raster data, and each pixel of the river raster data has at least the characteristics of river width, river channel direction, and off-shore distance.

[0124] The embodiments of the present disclosure have been described above. The above description is exemplary and not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of the terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technologies in the market, or to enable other ordinary technicians in the technical field to understand the embodiments disclosed herein.

Claims

1. A method for calculating multi-dimensional characteristics of rivers in remote sensing images, characterized in that including: extracting a river area and a river boundary line in a target image; obtaining a river center line based on the river area, and uniformly generating a plurality of center points along the river center line; for each center point, determining a corresponding river cross-section line; uniformly generating a plurality of cross-section points along each river cross-section line, where the cross-section points have at least a river width feature, a river channel direction feature, and an offshore distance feature. Among them, taking the length of the river cross-section line corresponding to the cross-section point as the river width feature value, calculating the river channel angle of the cross-section point as the river channel direction feature value, and calculating the distance between the cross-section point and the river boundary line as the offshore distance feature value; using a preset interpolation method to interpolate the feature values of the cross-section points into the river area to generate river raster data, and each pixel of the river raster data has at least a river width feature, a river channel direction feature, and an offshore distance feature.

2. The method according to claim 1, wherein The extracting a river area and a river boundary line in a target image includes: identifying a water body area in the target image as the river area, and converting the river area into planar vector data; performing preprocessing on the planar vector data, and extracting the river boundary line.

3. The method according to claim 2, wherein The performing preprocessing on the planar vector data and extracting the river boundary line includes: using a key bend retention geometric simplification algorithm to simplify the planar vector data; identifying and filling void areas or fracture areas in the planar vector data; using a polynomial approximation method based on an exponential kernel to smooth the planar vector data; extracting the river boundary line based on the planar vector data after the above preprocessing is completed.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: uniformly generating a plurality of boundary points along the river boundary line at a first preset step size.

5. The method according to claim 4, wherein Before performing the step of for each center point, determining a corresponding river cross-section line, the method further includes: obtaining the minimum distance d_min between the center point and all boundary points; connecting two adjacent center points to the center point, and determining the river cross-section direction corresponding to the center point according to the orthogonal direction of the connection line; Calculate the azimuth angle between the direction of the river channel cross-section and the preset direction .

6. The method according to claim 5, characterized in that, The for each center point, determining a corresponding river cross-section line includes: using the center point as a starting point, respectively extending lines to both sides along the corresponding river cross-section direction, and the length of the extension line is a preset multiple of d_min; retaining the line segment of the extension line within the river area as the river cross-section line corresponding to the center point.

7. The method according to claim 6, wherein The calculating the river channel angle of the cross-section point as the river channel direction feature value includes: The channel angle of the cross-section point is calculated using the following formula :[[]]END]] wherein, saa is the solar azimuth angle at the cross-section point, which is obtained from the metadata of the target image; is the azimuth angle corresponding to the river channel cross-section direction of the cross-section point.

8. The method according to claim 1, wherein The calculating the distance between the cross-section point and the river boundary line as the offshore distance feature value includes: obtaining two intersection points of the river cross-section line corresponding to the cross-section point and the river boundary line; respectively calculating the distances between the cross-section point and the two intersection points as the offshore distance feature value.

9. The method according to claim 1, wherein The using a preset interpolation method to interpolate the feature values of the cross-section points into the river area includes: using an inverse distance weighted interpolation method to interpolate the feature values of each cross-section point into the river area.

10. A multi-dimensional feature calculation system for rivers in remote sensing images, characterized in that, including: an extraction module configured to extract a river area and a river boundary line in a target image; a center point module configured to obtain a river center line based on the river area, and uniformly generate a plurality of center points along the river center line; A river cross-section line module, configured to determine a corresponding river cross-section line for each center point; A cross-section point module, configured to uniformly generate a plurality of cross-section points along each river cross-section line respectively, where the cross-section points have at least a river width feature, a river channel direction feature, and an offshore distance feature. Among them, the length of the river cross-section line corresponding to the cross-section point is used as the river width feature value, the river channel angle of the cross-section point is calculated as the river channel direction feature value, and the distance between the cross-section point and the river channel boundary line is calculated as the offshore distance feature value; An interpolation module, configured to interpolate the feature values of the cross-section points into the river area by using a preset interpolation method to generate river raster data, and each pixel of the river raster data has at least a river width feature, a river channel direction feature, and an offshore distance feature.

Images