Compressed Sensing and Multiple Image Fusion: An Information Theoretic Approach

In this paper, we propose an information theoreticapproach to fuse images compressed by compressed sensing (CS)techniques. The goal is to fuse multiple compressed imagesdirectly using measurements and reconstruct the final image onlyonce. Since the reconstruction is the most expensive step, it wouldbe a more economic method than separate reconstruction of eachimage. The proposed scheme is based on calculating the resultusing weighted average on the measurements of the inputs, whereweights are calculated by information theoretic functions. Thesimulation results show that the final images produced by ourmethod have higher quality than those produced by traditionalmethods, especially if the number of input images exceeds two.

Keywords:

compressed sensing , image fusion , information theory