An Adaptive Fusion of Noisy Images for Multi Modal Medical Image Application

Multimodal medical images lends to certain problems in the medical imaging fusion system, since the devices introduces noise during the process of image capturing. Because of the appearance of noises in the input images, it leads to trouble as introducing artifacts in the resulted image of fusing the degraded images. This paper proposes a fusion approach for noisy images, captured from two different modalities, that merge adaptively unique details of the images denoised by i) bilateral method ii) curvelet method iii) 2D modulated filter bank method. For the fusion of denoised images, weight may be calculated by residuals of the recommended filtering methods. Performance measures required to show the stability of the proposed method are effectively employed in comparing the results with other state-of-the-art fusion methods.