Image fusion of natural, satellite, and medical images using Undecimated Discrete Wavelet Transform and contrast visibility

The objective of image fusion is to combine multiple input images into a single composite image. Discrete Wavelet Transform (DWT) is widely used discrete mathematical techniques in analysis and synthesis of images. In this family, the Undecimated Discrete Wavelet Transform (UDWT) method works more efficiently in real-time systems in combining multifocus images into a single composite image. Existing DWT based methods for image fusion are suffering from undesirable side effects like blurring which reduce the quality of the output image. The discrete implementation of UDWT can be accomplished by using `à trous' (with holes) algorithm. In this paper, an efficient method for fusion of multifocus images based on UDWT and contrast visibility is presented. First, the images to be fused are convolved with a predefined kernel then the edge features are extracted and contrast visibility is calculated for the edge features. Finally, the fused image is obtained by merging all edge planes and the residual plane. Experimental results on several pairs of multifocus images verify that the proposed method is consistent and preserves more information compared to earlier methods like Principal Component Analysis (PCA), Discrete Wavelet Transform (DWT) and it is also compared with various image quality metrics such as Mutual information (MI), Spatial frequency (SF), and Edge-based image fusion metric (QAB/F).