Description

Multi sensor image fusion enhances the human visual perception and machine interpretation of the scene by integrating complementary and redundant information given by multi sensor data. In this paper, we proposed a multi sensor image fusion method that provides a high contrast fused image having no structural bias and which is more robust to different types of source images. These objectives are achieved through an intelligent ensemble of guided image filter, nonsubsampled shearlet transform, texture energy measures, and morphological operations. The proposed method is validated on medical, infrared-visible, and multi focus images. The qualitative and quantitative assessment proved the superiority of the proposed method compared to state of the art image fusion methods