Image fusion algorithms based On contrast in multiwavelet domain

王志慧,赵保军,沈庭芝

DOI: https://doi.org/10.3321/j.issn:1001-506X.2008.10.012

2008-01-01

Abstract:Aiming at the application of multiwavelet theory in image processing field,the definition of image directional contrast in multiwavelet domain is given.In addition,an adaptive weighted-average fusion algorithm and selective fusion one which are based on multiwavelet transform and directional contrast are proposed.This algorithm decomposes original images by multiwavelet transform,then corresponding directional contrast are calculated according to decomposing coefficients,and the fusion rule based on contrast is set up.The experiments using these algorithms to multi-spectral image are done.Moreover,the comparative results of evaluating fusion quality are listed.The results show that image fusion algorithms based on directional contrast defined by multiwavelet transform perform obviously better in reserving image details of the information and increasing the amount of information.