A Hybrid Image Fusion and Denoising Algorithm based on Multi-scale Transformation and Signal Sparse Representation

Abstract

In response to the problem of denoising in image fusion, the author proposes a hybrid image fusion and denoising algorithm based on multi-scale transformation (MLT) and signal sparse representation (SRS). A hybrid model is constructed for shear transformation, and the coefficients after MLT decomposition are thresholded. Sliding window technology and translation invariance are used to form sparse representation for image fusion, and SRS algorithm is used to remove noise from the source image. The experimental results show that the algorithm reduces the contrast and spectral information distortion of the fused image, displays high-quality visual fusion effects, maintains high PSNR values under different noise levels, can provide a more complete description of the features in the image, accurately judge the focus area, maintain the structural correlation of the image, and strengthen the description of fusion edges and details in the fused image. It has been proven that the methods of multi-scale transformation and sparse signal representation can fuse and denoise images.