Magnetic resonance and computer tomography image fusion using novel weight maps obtained by using median and guided filters

Abstract

This paper proposes a technique for medical picture fusion based on the guided image filter. It utilizes guided filtering to smooth images, using texture information as guidance for the filter. Weight maps of the detail images are created through pixel weight computation based on image statistics. Finally, the source images are combined using a weighted average combining approach. The effectiveness of the proposed method is evaluated in terms of multiple quantitative image quality assessment factors and compared with several state-of-the-art image fusion algorithms. Experimental results indicate that the suggested approach for image fusion is effective.