A fully automatic lake positioning method and system based on domestic optical remote sensing satellite images

Abstract

The present invention discloses a fully automatic lake positioning method and system for domestic optical remote sensing satellite images, including collecting publicly published time series lake data sets or self-built time series lake data sets to construct a lake image base database; calculating the lake area of ​​the remote sensing image to be corrected; searching the base database for an image whose lake area is closest to the lake area of ​​the image to be corrected as a reference image, and resampling its spatial resolution to be consistent with the image to be corrected; converting the reference image to the coordinate system of the lake image to be corrected, moving the lake image to be corrected with the reference image as the reference, selecting the translation parameters corresponding to the maximum overlapping area as the correction parameters for image correction, and realizing lake positioning. The present invention uses a method that matches the historical water body range to correct images, does not need to use ground control points, can automatically and accurately correct images, and has a prospect for commercial promotion.

Description

Technical Field

[0001] The invention belongs to the technical field of geometric positioning and remote sensing image processing, and specifically relates to a fully automatic lake positioning method and system for domestic optical remote sensing satellite images. Background Art

[0002] my country's high-resolution optical remote sensing satellites are developing rapidly. The number of satellites in the Gaofen and Ziyuan series is increasing. The earth observation capabilities of domestic remote sensing satellites have been greatly improved, and they have played an important role in land use, agricultural monitoring, ocean monitoring, weather forecasting, etc. With the vigorous development of domestic optical remote sensing satellites, the geometric accuracy of domestic optical high-resolution remote sensing satellites has also attracted attention. The geometric positioning accuracy of satellites directly affects subsequent data applications and is an important part of data preprocessing. In terms of lake monitoring, many studies focus on water color element parameter concentration, inherent optical properties, hydrological elements, environmental parameters, etc. In the process of obtaining time series products of these element parameters, accurate satellite geometric positioning information is crucial.

[0003] At present, in response to the problem of insufficient geometric accuracy of images from my country's high-resolution optical remote sensing satellites, many researchers have carried out corrections based on ground calibration fields. However, a large number of ground control points need to be manually selected as matching point pairs. This method is time-consuming and labor-intensive, and it is difficult to balance timeliness and accuracy. More efficient and high-precision geometric correction methods have attracted widespread attention. Lakes have obvious distinguishing features from urban, mountainous and other backgrounds in optical remote sensing satellite images, and can be easily and accurately extracted. Many researchers and organizations have publicly published long-term lake datasets, such as the ReaLSAT dataset and the Dynamic World dataset. These datasets cover lakes around the world, provide lake ranges accurate to the monthly level, and have rich contour and area information in the binary map of lake water bodies in the image, which makes it possible to automatically locate lakes. Summary of the invention

[0004] In view of the shortcomings of the existing technology, the present invention provides a fully automatic lake positioning method and system for domestic optical remote sensing satellite images, which uses a method that matches the historical water body range to correct the image, does not require the use of ground control points, and can automatically and accurately correct the image, thus having prospects for commercial promotion.

[0005] In order to achieve the above object, the present invention provides a fully automatic lake positioning method using domestic optical remote sensing satellite images, comprising the following steps:

[0006] Step 1, establishing a lake image base database of the remote sensing image to be corrected;

[0007] Step 2, extracting lakes in the remote sensing image to be corrected;

[0008] Step 2.1, using the normalized difference water index NDWI to extract the lakes in the remote sensing image to be corrected;

[0009] Step 2.2, smoothing the extracted lake contour to remove image noise;

[0010] Step 2.3, clip to obtain the lake range and calculate the lake area;

[0011] Step 3, converting the lake vector image in the base database into a raster image;

[0012] Step 4, searching the base database for an image whose lake area is closest to the lake area of ​​the image to be corrected as the reference image;

[0013] Step 5, resampling the spatial resolution of the reference image to be consistent with the image to be corrected;

[0014] Step 6, converting the reference image to the coordinate system of the lake image to be corrected extracted in step 2, moving the lake image to be corrected based on the reference image, and selecting the x- and y-axis translations corresponding to the maximum overlapping area as correction parameters;

[0015] Step 7: Use the correction parameters to correct the image to achieve lake positioning.

[0016] Moreover, in step 1, according to the longitude and latitude of the remote sensing image to be corrected, a publicly published long-time series lake data set is downloaded to establish a lake image base database, or lakes are extracted based on remote sensing images after precise geometric correction to establish a lake image base database. The images in the lake image base database are visually inspected, and the correctable broken spots caused by clouds, cloud shadows, and noise are deleted, and the water body extraction results that are obviously contaminated by clouds are removed to form a lake image base database containing longitude and latitude, time, lake vector range, and lake area information.

[0017] Moreover, the normalized water index in step 2.1 is:

[0018]

[0019] Where NDWI is the normalized water index, GREEN is the green light band of the image to be corrected, and NIR is the near infrared band of the image to be corrected.

[0020] Moreover, after the outlines of all lake patches extracted in step 2.1 are generated in step 2.2, the outlines containing pixels less than n% of the historical maximum area in the lake base database are filtered out, where n is a set threshold.

[0021] Moreover, in the step 2.3, the area of ​​each lake vector generated in step 2.2 is calculated, and the east-west width of the historical largest area vector in the lake image base database is calculated, each lake vector is buffered outward with the width value, and the circumscribed rectangle of the buffered figure is used as a mask to crop the remote sensing image to be corrected, so as to obtain the lake image to be corrected.

[0022] Moreover, the raster image in the base database in step 4 has an area attribute, and the lake area extracted from the image to be corrected is used to find the closest lake area in the base database, and the corresponding lake raster image is obtained as the reference image.

[0023] Moreover, in step 6, it is assumed that the minimum value of the x direction of the reference image converted to the coordinate system of the lake image to be corrected is x j min , the maximum value is x j max , the minimum value in the y direction is y j min , the maximum value is y j max , the range of the lake image to be corrected in the x direction is (x d min ,x d max ), the range of the lake image to be corrected in the y direction is (y d min ,y d max ), then the moving range of the lake image to be corrected in the positive x direction is (0,|x j max -x d min |), the moving range in the negative x direction is (0,|x dmax -x jmin ), the moving range in the positive y direction is (0,|y j max -y d min |), the moving range in the negative y direction is (0,|y dmax -y jmin |); Each time it moves, it moves in the x and y directions simultaneously. Each time it moves, the overlapping area between the lake image to be corrected and the reference image is calculated, and the x and y axis translation corresponding to the maximum overlapping area is selected as the correction parameter.

[0024] The present invention also provides a fully automatic lake positioning system for domestic optical remote sensing satellite images, which is used to implement the fully automatic lake positioning method for domestic optical remote sensing satellite images as described above.

[0025] Moreover, it includes a processor and a memory, the memory is used to store program instructions, and the processor is used to call the program instructions in the memory to execute the fully automatic lake positioning method of a domestic optical remote sensing satellite image as described above.

[0026] Alternatively, it includes a readable storage medium on which a computer program is stored, and when the computer program is executed, it implements the fully automatic lake positioning method of a domestic optical remote sensing satellite image as described above.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] The present invention adopts a method matching the historical water body range to correct the image, does not need to use ground control points, can automatically and accurately correct the image, and has a prospect for commercial promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 The figure is a flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings and embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] Example 1

[0033] This embodiment uses the geometrically offset image of the Danjiangkou Reservoir in Hubei Province on August 13, 2022 taken by the Ocean No. 1C satellite as research data, and further illustrates the technical solution of the present invention in conjunction with the accompanying drawings.

[0034] like Figure 1 As shown, the embodiment of the present invention provides a fully automatic lake positioning method using domestic optical remote sensing satellite images, comprising the following steps:

[0035] Step 1: Establish a lake image base database for the remote sensing image to be corrected.

[0036] According to the longitude and latitude of the remote sensing image to be corrected, download the publicly published long-time series lake data set, such as the Reservoir and Lake Surface Area Time Series Dataset (ReaLSAT) and Dynamic World, as the lake image base database; or use the normalized water body index to establish a lake image base database based on the water body range data extracted from Landsat-5 / 7 / 8, Sentinel-2 and other images after precise geometric correction. In this embodiment, a lake image base database is established by extracting lakes from Sentinel-2 images that have undergone precise geometric correction. The images in the lake image base database are visually inspected, and the correctable broken spots caused by clouds, cloud shadows, noise, etc. are deleted, and the water body extraction results that are obviously contaminated by clouds are removed (that is, water bodies with unconnected pixels, a large number of broken spots inside the lake vector range, and sudden changes in lake area at adjacent times are eliminated), forming a lake image base database containing longitude and latitude, time, lake vector range, and lake area information.

[0037] Step 2: Extract lakes from the remote sensing image to be corrected.

[0038] Step 2.1, use the normalized water index to extract the lake in the remote sensing image to be corrected.

[0039] The normalized water index is:

[0040]

[0041] Where NDWI is the normalized water index, GREEN is the green light band of the image to be corrected, and NIR is the near infrared band of the image to be corrected.

[0042] In this embodiment, the value of the threshold δ is set to 0. When the NDWI of a pixel point is greater than 0, it indicates that it is a lake and is extracted.

[0043] Step 2.2, smooth the extracted lake contour to remove image noise.

[0044] Call the findContours function in the opencv library to generate the contours of all lake patches extracted in step 2.1, and filter out contours containing pixels that are less than 10% of the historical maximum area in the lake basement database.

[0045] Step 2.3, clip to obtain the lake range and calculate the lake area.

[0046] Calculate the area of ​​each lake vector (patch outline) generated in step 2.2 in the ArcGIS field calculator, and calculate the east-west width of the historical largest area vector in the lake image base database. Buffer each lake vector outward with this width value, and use the circumscribed rectangle of the buffered figure as a mask to crop the remote sensing image to be corrected to obtain the lake image to be corrected.

[0047] Step 3, convert the lake vectors in the base database into raster data.

[0048] The public datasets collected in step 1 and the datasets established by extracting lakes from remote sensing images after precise geometric correction are both raster images. In order to obtain the range and area information of the lakes when building the database, the raster images were converted into vector images. In order to subsequently convert the reference images to the coordinate system of the lake image to be corrected and calculate the translation parameters, the vector images in the base database need to be converted into raster images. The RasterizeLayer function of the GDAL library is used to convert the lake vector images in the base database into raster images.

[0049] Step 4, searching the base database for an image whose lake area is closest to the lake area of ​​the image to be corrected as the reference image.

[0050] The raster images in the base database have area attributes. The lake area extracted from the image to be corrected is used to find the closest lake area in the database, and the corresponding lake raster image is obtained as the reference image.

[0051] Step 5: resample the spatial resolution of the reference image to be consistent with the image to be corrected.

[0052] The image resolution in the base database is higher than that of the image to be corrected. The ReprojectImage function of the GDAL library is used to resample the base image to make its spatial resolution consistent with that of the image to be corrected.

[0053] Step 6, convert the reference image to the coordinate system of the lake image to be corrected extracted in step 2, move the lake image to be corrected based on the reference image, and select the x and y axis translation amounts corresponding to the maximum overlapping area as correction parameters.

[0054] Assume that the minimum value of the x direction of the reference image converted to the coordinate system of the lake image to be corrected is x j min , the maximum value is x j max , the minimum value in the y direction is y j min , the maximum value is y j max , the range of the lake image to be corrected in the x direction is (x d min ,x d max ), the range of the lake image to be corrected in the y direction is (y d min ,yd max ), then the moving range of the lake image to be corrected in the positive x direction is (0,|x j max -x d min |), the moving range in the negative x direction is (0,|x dmax -x jmin |), the moving range in the positive y direction is (0,|y j max -y d min |), the moving range in the negative y direction is (0,|y dmax -y jmin |). Each time it moves, it moves in the x and y directions simultaneously. Each time it moves, the overlapping area between the lake image to be corrected and the reference image is calculated, and the x and y axis translations corresponding to the maximum overlapping area are selected as correction parameters. In this embodiment, the x and y axis translations corresponding to the maximum overlapping area are (-223, -1195).

[0055] Step 7: Use the correction parameters to correct the image to achieve lake positioning.

[0056] The GDAL library is used to read the coordinates of the upper left corner of the image to be corrected, and then the GeoTransform parameters are rewritten with the correction parameters (-223, -1195) to achieve lake positioning.

[0057] Experimental verification

[0058] The synchronous sentinel image of the image to be corrected was selected, and the correction accuracy was quantitatively evaluated based on the change of the centroid of the water body in the image to be corrected before and after correction and the centroid of the water body extracted from the sentinel image. The lake water body in the sentinel image was extracted using the normalized water body index, and the noise was removed and cropped as verification data. The raster data was converted into vector data using the FPolygonize function in the GDAL library.

[0059] The computational geometry tools in ArcGIS were used to calculate the vector centroid coordinates of the water bodies extracted from the synchronous sentinel images and the images to be corrected before and after correction, and the correction accuracy was tested based on the distance index d:

[0060]

[0061] Where x1 is the x-direction projection coordinate of the water body centroid extracted from the sentinel image, x2 is the x-direction projection coordinate of the water body centroid extracted from the image to be corrected before / after correction, y1 is the y-direction projection coordinate of the water body centroid extracted from the sentinel image, and y2 is the y-direction projection coordinate of the water body centroid extracted from the image to be corrected before / after correction.

[0062] Calculate the distance between the centroid of the water body of the image to be corrected before and after correction and the centroid of the water body of the sentinel image respectively. The distance after correction should be smaller than the distance before correction, and the closer to 0, the better.

[0063] In this embodiment, the centroid coordinates of the Danjiangkou Reservoir extracted from the Sentinel-2 image on August 13, 2022 are (537707, 3615208), and the centroid coordinates of the water body extracted from the Ocean-1C satellite on August 13, 2022 are (538090, 3617320). The centroid coordinates after correction are (537867, 3616125), d1 = 2146.45 before correction, and d2 = 930.85 after correction. The distance between the centroid of the water body and the reference centroid after correction is less than the distance between the centroid of the water body and the reference centroid before correction.

[0064] Example 2

[0065] Based on the same inventive concept, the present invention also provides a fully automatic lake positioning system for domestic optical remote sensing satellite images, including a processor and a memory, the memory is used to store program instructions, and the processor is used to call the program instructions in the memory to execute the fully automatic lake positioning method for domestic optical remote sensing satellite images as described above.

[0066] Example 3

[0067] Based on the same inventive concept, the present invention also provides a fully automatic lake positioning system for domestic optical remote sensing satellite images, including a readable storage medium, on which a computer program is stored. When the computer program is executed, it implements the fully automatic lake positioning method for domestic optical remote sensing satellite images as described above.

[0068] In specific implementation, the method proposed in the technical solution of the present invention can be implemented by technical personnel in this field using computer software technology to realize automatic operation process. System devices for implementing the method, such as computer-readable storage media storing the corresponding computer program of the technical solution of the present invention and computer equipment running the corresponding computer program, should also be within the scope of protection of the present invention.

[0069] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific implementation cases described, or replace them with similar methods, without departing from the spirit of the present invention or exceeding the scope defined by the appended claims.

Claims

1. A fully automatic lake positioning method using domestic optical remote sensing satellite images, characterized in that: The following steps are involved: Step 1, establishing a lake image base database of the remote sensing image to be corrected; Step 2, extracting lakes in the remote sensing image to be corrected; Step 2.1, using the normalized difference water index NDWI to extract the lakes in the remote sensing image to be corrected; Step 2.2, smoothing the extracted lake contour to remove image noise; Step 2.3, clip to obtain the lake range and calculate the lake area; Step 3, converting the lake vector image in the base database into a raster image; Step 4, searching the base database for an image whose lake area is closest to the lake area of ​​the image to be corrected as the reference image; Step 5, resampling the spatial resolution of the reference image to be consistent with the image to be corrected; Step 6, converting the reference image to the coordinate system of the lake image to be corrected extracted in step 2, moving the lake image to be corrected based on the reference image, and selecting the x- and y-axis translations corresponding to the maximum overlapping area as correction parameters; Step 7: Use the correction parameters to correct the image to achieve lake positioning.

2. The fully automatic lake positioning method of a domestic optical remote sensing satellite image as claimed in claim 1, characterized in that: In step 1, according to the longitude and latitude of the remote sensing image to be corrected, the publicly published long-term lake dataset is downloaded to establish a lake image base database, or the lake is extracted by the remote sensing image after precise geometric correction to establish a lake image base database; Visually inspect the images in the lake image base database, delete the correctable broken patches caused by clouds, cloud shadows, and noise, remove the water body extraction results contaminated by clouds, and form a lake image base database containing latitude and longitude, time, lake vector range, and lake area information.

3. The fully automatic lake positioning method of a domestic optical remote sensing satellite image as claimed in claim 1, characterized in that: The normalized water index in step 2.1 is: Where NDWI is the normalized water index, GREEN is the green light band of the image to be corrected, and NIR is the near infrared band of the image to be corrected.

4. The fully automatic lake positioning method of a domestic optical remote sensing satellite image as claimed in claim 1, characterized in that: After generating the outlines of all lake patches extracted in step 2.1 in step 2.2, the outlines containing pixels less than n% of the historical maximum area in the lake base database are filtered out, where n is the set threshold.

5. The fully automatic lake positioning method of a domestic optical remote sensing satellite image as claimed in claim 1, characterized in that: In step 2.3, the area of ​​each lake vector generated in step 2.2 is calculated, and the east-west width of the historical largest area vector in the lake image base database is calculated. Each lake vector is buffered outward with this width value, and the circumscribed rectangle of the buffered figure is used as a mask to crop the remote sensing image to be corrected to obtain the lake image to be corrected.

6. The fully automatic lake positioning method of a domestic optical remote sensing satellite image as claimed in claim 1, characterized in that: In step 4, the raster image in the base database has an area attribute. The lake area extracted from the image to be corrected is used to find the closest lake area in the base database, and the corresponding lake raster image is obtained as the reference image.

7. The fully automatic lake positioning method of domestic optical remote sensing satellite images as claimed in claim 1, characterized in that: In step 6, it is assumed that the minimum value in the x direction of the reference image converted to the coordinate system of the lake image to be corrected is x jmin , the maximum value is x jmax , the minimum value in the y direction is y jmin , the maximum value is y jmax , the range of the lake image to be corrected in the x direction is (x dmin ,x dmax ), the range of the lake image to be corrected in the y direction is (y dmin ,y dmax ), then the moving range of the lake image to be corrected in the positive x direction is (0,|x jmax -x dmin |), the moving range in the negative x direction is (0,x dmax -x jmin ), the moving range in the positive y direction is (0,|y jmax -y dmin |), the moving range in the negative y direction is (0,|y dmax -y jmin |); Each time it moves, it moves in the x and y directions simultaneously. Each time it moves, the overlapping area between the lake image to be corrected and the reference image is calculated, and the x and y axis translation corresponding to the maximum overlapping area is selected as the correction parameter.

8. A fully automatic lake positioning system based on domestic optical remote sensing satellite images, characterized in that: It includes a processor and a memory, the memory is used to store program instructions, and the processor is used to call the program instructions in the memory to execute a fully automatic lake positioning method for domestic optical remote sensing satellite images as described in any one of claims 1-7.

9. A fully automatic lake positioning system based on domestic optical remote sensing satellite images, characterized in that: It includes a readable storage medium, on which a computer program is stored. When the computer program is executed, a fully automatic lake positioning method for domestic optical remote sensing satellite images as described in any one of claims 1 to 7 is implemented.

Images