Multiband image fusion method based on multiscale transformation and possibility theory

Abstract

The invention relates to an image fusion method, in particular relates to a multiband image fusion method, and more particularly relates to a multiband image fusion method based on multiscale transformation and possibility theory. The method comprises the following steps of respectively carrying out support degree transformation on multiple bands of images to obtain low-frequency images and support-degree sequence images; decomposing a last layer of low-frequency component image with most scattered gray value by utilizing a quadtree method; respectively decomposing the last layer of low-frequency component image of other bands of images by adopting the quadtree decomposition result as an evidence; respectively fusing block images corresponding to multiband low-frequency components by adopting an extraction mode to obtain low-frequency fusion block images; splicing the fused block images to obtain a low-frequency fusion image; and executing the support-degree inverting transformation on the low-frequency fusion image and the support-degree sequence image synthesized by utilizing a maximal method to obtain a final synthesized image. According to the image fusion method, the marginal strength, the contrast ratio and the entropy of the synthesized image can be effectively enhanced; meanwhile, the running time is greatly shortened.

Description

technical field [0001] The invention relates to an image fusion method, in particular to a multi-band image fusion method, specifically a multi-band image fusion method based on multi-scale transformation and possibility theory. Background technique [0002] Multi-band imaging is an important means of disaster monitoring, fault diagnosis, industrial monitoring and deep space exploration. The purpose is to use the difference and complementarity of detection results of different bands to obtain a more comprehensive and accurate understanding of the scene. For example, the ability of infrared long-wave detection to penetrate smoke is stronger than that of medium-wave and visible light; the ability of infrared medium-wave detection under high humidity conditions is better than other wave bands; and the target edge and texture of visible light images are significantly better than infrared images. . However, although multi-band detection can obtain rich scene information, it usua...

AI Technical Summary

Problems solved by technology

[0006] The present invention aims to solve the problem of reducing the difference between original images caused by sequential fusion in multi-band image fusion, the problem of poor detail and texture information of feature-level fusion effect, and the limited and ubiquitous application range of quadtree decomposition fusion. Fusion algorithm is relatively complicated, and a multi-band image fusion method based on multi-scale transformation and possibility theory is provided.

Images