Hybrid multi-scale analysis-based infrared and visible light image fusion algorithm

Abstract

The present invention discloses a hybrid multi-scale analysis-based infrared and visible light image fusion algorithm. The algorithm comprises the following steps that: step 1, NSCT (nonsubsampled contourlet transform) decomposition is performed on infrared and visible images to obtain low-frequency subbands and high-frequency subbands; step 2, static wavelet transformation is performed on the low-frequency subbands so as to obtain a low-frequency subband and three high-frequency subbands, and a local energy and maximum absolute value selection-combined method and a compressed sensing theory are adopted to fuse the low-frequency subband and the three high-frequency subbands; step 3, the definition of an image to be fused is judged, and the enhancement layers of an LSCN are selected according to judging criteria; step 4, the topmost layer high-frequency subband is fused according to a fusion rule of selecting a largest absolute value, and the other subbands are fused through adopting an improved PCNN model; and step 5, inverse NSCT is performed on fusion results, so that a final fused image is obtained. The fused image obtained by using the method of the invention has prominent edges and high contrast; a target in the fused image is prominent; and the indexes of the algorithm such as average gradient and spatial frequency are higher than those of algorithms in the prior art.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an infrared and visible light image fusion algorithm based on mixed multi-scale analysis. Background technique [0002] The image fusion method based on wavelet transform is a classic fusion algorithm, but wavelet can only represent isotropic objects, and wavelet is not an ideal representation tool for features such as image center line and edge. Contourlet is widely used in image fusion. Through the multi-scale and multi-directional decomposition of images, Contourlet can capture the details of the image very well, making up for the shortcomings of wavelet. However, due to the down-sampling operation used in the Contourlet transform, it does not have translation invariance, and it is easy to produce pseudo-Gibbs phenomenon in image processing. [0003] The non-subsampled Contourlet transform (NSCT) proposed by A.L. Cunha has translation invariance, which ca...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the problem solved by the present invention is how to solve the problems of low contrast and insufficient edge information retention in the fusion of infrared and visible light images.

Images