A Compressed Sensing and Porous 9-7 Wavelet Transform-based Image Fusion Algorithm

Image fusion is ubiquitous in the area of image processing. However, after decomposing an original image for feature extraction in a fusion process, its most low-frequency sub-bands are redundant, while its vital high-frequency ones are not adequately extracted. Moreover, in context of multi-image fusion, Pseudo-Gibbs effect is very difficult to eliminate. To address the above issues, this study proposes a compressed sensing and porous 9-7 wavelet transform-based image fusion algorithm (CPIF). Its main ideal is two-fold, a) reducing the fusion time in a sparse feature representation space, and b) releasing the storage space especially in multi-type image fusion. Results on medical diagnostic images and multi-images in practical industry applications indicate that a CPIF algorithm is efficient and robust in addressing the task of image fusion when compared with five state-of-the-art wavelet-transform methods.