Feature fine matching method and application based on minimum spanning tree

Abstract

The invention belongs to the technical field of digital image processing, and discloses a feature fine matching method and application based on a minimum spanning tree, and the method comprises the steps: converting an image into an HSL space, and carrying out the convolution of an L channel through a multi-scale Gaussian function; correcting the illumination component of the image to enhance a brightness non-uniform region in the image; equalizing brightness components among the three-line array CCD images after dodging; adopting surf feature matching for the preprocessed three-line array CCD dodging image to serve as rough matching feature points; calculating an affine transformation matrix by roughly matching the affine transformation relation of the feature point pair; generating an image minimum spanning tree according to the color in the image and the distance information, and taking the image minimum spanning tree as a weight to constrain a least square matching algorithm; and calculating a matching window correlation coefficient, continuously iterating the search window, and outputting by taking the center point coordinate as a matching point pair. The sparse matching provided by the invention has good registration precision, the average matching precision of the algorithm is 97.21%, and the average matching time is 6.3443 s.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, in particular to a feature fine matching method based on a minimum spanning tree and its application. Background technique [0002] At present, remote sensing images (Remote Sensing Image) can truly and clearly display the position, shape, size, color and other information of objects on the earth's surface, and become an important data source for basic geographic data collection and update. my country has successfully reflected three surveying and mapping satellites, "Tianhui No. 1" (TH-1) and "Ziyuan No. 3" (ZY-3), and has made significant progress in the research and development of high-resolution remote sensing satellite systems. Its spatial resolution Up to 2m. Sparse and dense matching of three-line array CCD images is an application requirement in geographic information collection and map surveying and mapping, and is also the basis for remote sensing surveying and mapping ...

AI Technical Summary

Problems solved by technology

[0005] (1) In the existing technology, the data volume of remote sensing images is large, and the images include complex texture areas such as cities and mountain villages, as well as low-texture and repeated-texture areas such as lakes, mountains, and terraced fields. The feature point matching algorithm is directly used in the matching Accuracy and matching time are less effective

[0006] (2) In the prior art, the image brightness is uneven and the brightness inconsistency between images is poor. The correlation measurement of the image matching window is based on the image grayscale information. The brightness of shadows in the image is uneven and the brightness of the same group of images is inconsistent. Directly affect the matching accuracy of subsequent image feature matching and correlation calculation

[0007] (3) In the prior art, in the least squares matching algorithm, the region window correlation is used as the feature fine matching, which lacks the globality of the matching window

In the evaluation criteria of brightness average, standard deviation, and information entropy in the five major terrain areas of remote sensing images such as cities, mountains, lakes, mountain villages and terraces Good; in order to solve the problem that the least squares local matching is not global, for the satellite three-line array CCD remote sensing image, with the strategy from coarse to fine, a feature fine matching algorithm based on the minimum spanning tree is proposed for image sparse matching process

Images