(IJCSIS) International Journal of Computer Science and Information Security, Vol. 14, No. 7, July 2016 A novel approach for multi-modality image fusion with conjugation of DWT and RT using region consistency Keyur N. Brahmbhatt Department of Computer Engineering, Research Scholar, Changa, India Dr. Ramji M. Makwana Department of Information Technology, Professor, VVP College, Rajkot, India Abstract—Complementary multi-modality image information of more than one image when joined together and create single new informative image, this is known as image fusion. Main purpose of it is to reduce ambiguity and idleness in a resultant image by enhancing relevant details specific to any task or any application. In medical imaging images comes from various input sources which has different detailed information. Thus there will be an interesting task to perform merging operation on registered multimodality images. The image fusion is very valuable in medical analysis. Here in our research paper, the fusion process has been done in two transform named discrete wavelet transform (DWT) & Ripplet transform (RT). Region consistency check and Maximum selection fusion rules has been used. Implementation task has been performed on Computed Tomography and Magnetic Resonance Imaging images. Evaluating and doing comparative study of fusing methods, measuring parameters are used. Implementation of method shows that, our suggested method exhibits a fine results in area of medical imaging, because our method provides arbitery scaling due to RT and good local features due to DWT. Keywords-: Image Fusion, DWT (Discrete Wavelet Transform), Ripplet Transform I. INTRODUCTION A procedure of conjugating two or more multimodality images to single informative image which is useful by any medical imaging application called multimodality image fusion [18].The research work into this direction can be put into following categories: • Primitive Fusion scheme • Discrete Wavelet Transform Fusion scheme • Contourlet Transform Fusion scheme • Ripplet Transform Fusion scheme Primitive fusion scheme contain spatial domain methods for example PCA [11], [16], averaging [16], Maximum Selection [16] etc. A weakness of these methods is it does not offer directional singularity [10]. Thus frequency domain has been used with area of image fusion, who delivers directional singularity. Discrete Wavelet Transform (DWT) is offering horizontal as well as vertical and diagonal directionality, which is limited. Contourlet transform has been introduced to enhance directionality. Contourlet transform (CT) offers C 2 directional singularity. The CT provides virtuous result with curves [12].The RT is having high dimensional simplification of the transform known as Curvelet Transform (CVT) which is accomplished presenting images at dissimilar levels and dissimilar directions. CVT utilizes a parabolic scaling law for achieving great anisotropic directionality. Here, the anisotropic property ensures to resolve two directional singularities along C 2 curves. While, RT gives the new frame structure with sparse illustration for any images without continuity along C d curves [4]. II. DISCRETE WAVELET TRANSFORM A decomposition of spatial frequency that offers a supple multi resolution study of an image is called Discrete Wavelet Transform (DWT)[3],[4], [16]. 2D DWT desires separate filter & down sampling in a horizontal direction and a vertical direction. This artifact 4 sub-bands in which it representing a horizontal frequency first then a vertical frequency second at each scale. This creates high (H) – high (H), high (H) - low (L), low (L) – high (H) and low (L) -low (L) image sub-bands. By recursively performing the similar structure to the low (L) – low (L) or any sub-band multi-resolution decomposition will be accomplished. Figure 1: 2- level DWT [9] 357 https://sites.google.com/site/ijcsis/ ISSN 1947-5500