Performance Comparision Of Wavelet Transforms In Multimodal Biomedical Image Fusion

Authors

  • Venkatesan.B, Yuvaraj.K , Swarga.S.R, Sumithaa.S,Sugitha.S

Abstract

Image fusion is a process of combining two or more images into a single image to have all the source image information together and to provide a more detailed representation of the image for further processing of any applications. In recent years, image fusion is growing rapidly along with the advancement in imaging technology. Especially in the medical field, the imaging modalities are getting advanced in recent time which leads to the development of more and more algorithms for fusion. It is needed to have an effective and feasible algorithm for the clinical images. As time-frequency analysis method, wavelets are widely applied in fusion algorithms. The performance difference in the wavelet family constructs the difference evaluation in the fusion results. It is proposed to have an analytical approach of different wavelets in brain image fusion to find the better family for wavelet based fusion algorithms. The comparison results have shown that the higher order wavelet in each family performs better in the time-frequency based image fusion algorithm. Various works of literature are being studied and it is observed that most fusion is based on MRI and CT. Because MRI images provide better soft tissue definition and also higher spatial resolution, whereas CT images give importance to the three-dimension imaging which has short scan time and high imaging resolutions. About 55 Images of CT and MRI are taken from various sources and wavelet transform is performed on them by using suitable filters.  The fusion performance is evaluated based on Standard Deviation (SD) and Entropy (Epy).

Published

2020-11-01

Issue

Section

Articles