Change detection method for high-resolution remote sensing images based on local invariant features

Abstract

The invention discloses a high-resolution remote sensing image change detection method based on local invariant features. On the basis of twice cross-blocking of the image pair to be detected, LBP and SURF feature descriptors are used to analyze the similarity of the image block pair respectively. Confidence, based on the similarity confidence to judge the change nature of 1 / 4 image blocks, and use the method of morphological region growth to deal with the block effect formed by changing and non-changing 1 / 4 image blocks. The present invention divides the processing twice, analyzes the texture feature according to the image block, and judges the image change property according to the 1 / 4 image block, which improves the accuracy of the description of the image block and the precision of the analysis of the image block; The characteristics of variability and light insensitivity make the similarity judgment of image blocks more accurate; through morphological growth, the complexity of region growth threshold selection is reduced, and the processing of discontinuous small regions such as holes is avoided.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, belongs to multi-temporal remote sensing image change detection based on local invariant features, and in particular relates to a high-resolution remote sensing image change detection method based on LBP and SURF features. Background technique [0002] Remote sensing images have the characteristics of large coverage and intuitive reflection of the surface. Multi-temporal remote sensing image change detection is widely used in land monitoring, environmental monitoring, disaster monitoring, and urban planning. [0003] Remote sensing image change detection methods are mainly divided into pixel-based and object-based methods, and the features used mainly include grayscale and texture. The object-based method needs to segment the image to extract the analysis object, but there is a problem of segmentation scale in image segmentation, and there is a problem of feature selection in the...

AI Technical Summary

Problems solved by technology

[0005] In recent years, the spatial resolution of remote sensing images has become higher and higher, up to the decimeter level, and the geometric and texture details of objects in high-resolution remote sensing images are better presented in the image, but the grayscale of the image is affected by climate, Factors such as light intensity affect lower-resolution remote sensing images greatly, which brings challenges to high-resolution remote sensing image change detection

Images