A Edge Correction Method Applicable to Multiple Remote Sensing Images in Large Areas

Abstract

The invention relates to an edge correction method suitable for multiple remote sensing images in a large area. (1) Automatically match points of the same name in adjacent areas of all images to obtain points of the same name; (2) Use existing points of the same name to construct a triangulation network; (3) For each point with the same name, calculate the reference object coordinates of the point with the same name; (4) Calculate the affine transformation parameters of each image; (5) For each point with the same name, calculate the reference object coordinates of the point on different images and the corresponding Affine transformation parameters are used to calculate the new observed object-space coordinates of the point with the same name on the corresponding image; (6) Repeat (3) to (5) until the change of the reference object-space coordinates is less than a certain threshold or no longer changes; (7) Correct the corresponding image using the affine transformation parameters of the last iteration. The method proposed by the present invention can perform edge correction on multiple images in a large area at the same time; the correction accuracy is higher; and it avoids the problems of uneven deformation and excessive image distortion caused by traditional correction methods.

Description

technical field [0001] The invention relates to a border correction method suitable for multiple remote sensing images in a large area, belongs to the technical field of earth observation, and is a remote sensing image processing technology. Background technique [0002] In the application process of remote sensing images, according to the application scenarios and scope requirements, it is often necessary to use multiple remote sensing image products from different sources that have been processed and only have coordinate information to cover a certain range of target areas for work. At this time, these remote sensing image products that already have coordinate information often have poor mutual edge accuracy, which has a considerable impact on the application of data. [0003] To solve this problem, the traditional image edge correction algorithm adopts 1) using one of them as a reference to perform polynomial correction on the other image as a whole; 2) using energy The ...

AI Technical Summary

Problems solved by technology

[0004] 1. The traditional method can only deal with the edge correction between two images, and cannot process the edge correction of multiple remote sensing images in a large area;

[0005] 2. When one image is used as a reference to correct the other image as a whole, the part of the corrected image outside the overlapping area will have a large position shift, which will affect the relative accuracy;

[0006] 3. The secondary correction using the reference image is limited by whether the reference image can be completely covered, and the result of the secondary correction cannot guarantee that the edge accuracy between the two images meets the use requirements;

[0007] 4. Use the polynomial correction method to correct the overlapping area of ​​the two images. When the error is large, the use of low-order polynomials cannot guarantee the effect of edge correction, and the use of high-order polynomials will often cause excessive stretching of the corrected image ;

[0008] 5. When using the thin plate spline method to correct the edges between images with large errors, if the geometric deviation exceeds the specified tolerance, the correction result will often be deformed unevenly, and the excessive distortion of the image will affect the use of the situation

Images