Fusion of medical image using STSVD

Abstract

The process of uniting medical images which are taken from different types of images to make them as one image is a Medical Image Fusion. This is performed to increase the image information content and also to reduce the randomness and redundancy which is used for clinical applicability. In this paper a new method called Shearlet Transform (ST) is applied on image by using the Singular Value Decomposition (SVD) to improve the information content of the images. Here two different images Positron Emission Tomography (PET) and Magnetic Resonance Imaging (MRI) are taken for fusing. Initially the ST is applied on the two input images, then for low frequency coefficients the SVD method is applied for fusing purpose and for high frequency coefficients different method is applied. Then fuse the low and high frequency coefficients. Then the Inverse Shearlet Transform (IST) is applied to rebuild the fused image. To carry out the experiments three benchmark images are used and are compared with the progressive techniques. The results show that the proposed method exceeds many progressive techniques.