A Variational Fusion Method of Remote Sensing Image Based on Structure-Texture Decomposition

Abstract

The invention discloses a multispectral and panchromatic image variational fusion method based on structure-texture decomposition. Based on the similarity of structure components and texture components between PAN images and HRMS images in the field of remote sensing, the global information and local space are respectively analyzed. Details (high-level information) are preserved. Our variational model is proposed to ensure the structural similarity between PAN images and HRMS images by gradient sparsity, and to describe their texture similarity according to low-level attributes. The model is then solved by the method of alternating direction multipliers, resulting in multispectral images with high resolution. Experiments show that: the method proposed by the present invention takes into account the similarity of structural components and texture components between PAN images and HRMS images, and uses different simulated data sets and real data sets for fusion experiments. Compared with the comparison method, our fused image The indicators of are the closest to the ideal value, and visually the closest to the reference image.

Description

technical field [0001] The invention belongs to the field of image fusion, and relates to a multi-spectral and panchromatic image variational fusion method based on structure-texture (cartoon-texture) decomposition, which is suitable for fusion of low-resolution multi-spectral images and high-resolution panchromatic images Application scenarios of high-resolution multispectral imagery. Background technique [0002] In the field of remote sensing, the data of earth images acquired by satellites has important application value, such as land cover classification, disaster detection and vegetation mapping. Due to the limitation of satellite storage and transmission burden, different sensors are usually used to obtain the multispectral image (LRMS) with rich spectral information but low spatial resolution and the panchromatic image (PAN) with single spectral information but high spatial resolution of the same picture respectively. ), and by fusing them to obtain a multi-spectral...

AI Technical Summary

Problems solved by technology

However, the gradient operator can only obtain the first-order edge information of the image, so the high-order spatial information of the panchromatic image has not been effectively utilized in the fusion process.

Images