Least square matching method based on adaptive object-side element

Abstract

The invention belongs to the technical field of remote sensing image processing image matching, and discloses a least square matching method based on adaptive object-side elements, and the method comprises the steps: obtaining an initial object-side three-dimensional point coordinate through the spatial forward intersection and collinear equation projection relation of an initial matching image pair, and taking the initial object-side three-dimensional point coordinate as an object-side element main point; determining the size of a matched object aspect element and the position of a to-be-matched pixel in the image window according to the correlation of each pixel in the image window obtained by projecting the object-object element to the to-be-matched image by taking the object-object element main point as a reference, and establishing an error equation according to a spatial geometric projection relationship; and obtaining radiation distortion and geometric distortion correction values through least square iterative optimization, and searching for an optimal matching point position. The method can improve the matching precision and the matching success rate.

Description

technical field [0001] The invention relates to the technical field of remote sensing image processing and image matching, in particular to a least squares matching method based on adaptive object aspect elements. Background technique [0002] In recent years, the matching optimization method based on object aspect elements has been widely used to improve the accuracy of single-point matching. Its basic idea is: 1) use the initial matching to calculate the object point coordinates through forward intersection; 2) take the initial object point coordinates as the center Create a surface element in the object space; 3) Project the object points on the object surface element onto the reference image and the search image respectively according to the projection relationship; 4) Establish the error equation through the geometric projection relationship and use the least squares adjustment to calculate the optimal Match pixels and final object point coordinates. According to the s...

AI Technical Summary

Problems solved by technology

However, the tangent plane only exists on the surface of the continuous object, so the optimization method based on the object aspect element will have mismatching or low matching accuracy on the edge of the surface discontinuity, such as the building edge in the urban remote sensing image.

Moreover, for the least squares adjustment calculation, the initial parameters set will directly affect the matching results

When the difference between the initial parameters and the final result is large, it is likely that the solution cannot be solved because the correction value does not converge

Images