Multimodal medical image fusion using laplacian redecomposition

Abstract

Fusion of multi-modality images gives more sophisticated and complete information in video surveillance, remote sensing, and medical diagnosis. The present paper delivers a contemporary fusion method using laplacian redecomposition. The area of medical image fusion with different modalities has achieved more progress in medical field. Existing methods suffer from noise, blurring, and color distortion. These complications are addressed in the present method. The authors propose laplacian redecomposition method after enhancement to get redundant and complementary information. The proposed method has two alterations. First, we give the concept of overlap and non-overlap domains. Overlap and non-overlap domains are used for fusing redundant and complementary information respectively. Second, decision graph scheme is presented to get low frequency sub band images, complementary and redundant information. Global decision graph is used for reconstructing HF sub band images. Lastly, by taking inverse laplacian operation fused image is obtained. In the experiments for comparison, a few pairs of medical images have been taken to prove the strength of the suggested technique. The experimental results demonstrate that, the proposed technique accomplish aggressive performance qualitatively and quantitatively.