A categorization of multiscale-decomposition-based image fusion schemes with a performance study for a digital camera application

The objective of image fusion is to combine information from multiple image s of the same scene. The result of image fusion is a single image which is more suitable for human and machine perception or further image-processing tasks. In this paper, a generic image fusion framework based on multiscale decomposition is studied. This framework provides freedom to choose differ ent multiscale decomposition methods and different fusion rules. The framew ork includes ail of the existing multiscale-decomposition-based fusion appr oaches we found in the literature which did not assume a statistical model for the sour ce images. Different image fusion approaches are investigated based on this framework,. Some evaluation measures are suggested and applie d to compare the performance of these fusion schemes for a digital camera a pplication. The comparisons indicate that our-framework includes some new a pproaches which outperform the existing approaches Sol the cases Mie consid er.