Detail-Enhanced Multimodal Medical Image Fusion

Many traditional medical image fusion methods cannot well preserve details of source images in the fused image. Aiming at handling this issue, a novel medical image fusion scheme based on gain control is developed by nonsubsampled contour let transform (NSCT). To enhance image details, the proposed method applies power law transformation to tune coefficients of each decomposed sub band, and adjusts the strength of the sub band signals by smooth gain control. Finally, the fused image with more details is constructed by the inverse NSCT. Experimental results demonstrate that our method can produce better fusion performance than some existing methods by the comparison of visual perception and objective quality assessment.