Efficient image integration technique: mismatch of subjective and objective analysis

Abstract

The paper intends to extract features with high visual quality during image fusion. An improved image fusion algorithm is presented, aiming to preserve necessary features and relevant information, simultaneously focusing on highly enhanced visual results. The resultant fused image obtained is superior to other images obtained from other stateof-the-art techniques with respect to visual quality. A novel and improved image fusion algorithm is presented, which outperforms the other eight state-of-the-art techniques of image fusion. The proposed methodology includes a combinational approach of anisotropic diffusion and Laplacian pyramid leading to image enhancement. The paper also reveals a contradiction of the visual results with respect to objective results.