A Wavelet and Contourlet Based Hybrid Fusion for Medical Images Using Laplacian Pyramid and Directional Filter Banks with Weighted Average Method

摘要

Hybrid image fusion techniques enables the researchers to utilize the advantages of different fusion techniques, as a result, a more enhanced, accurate, and reliable image is obtained as an output with more anatomical and functional information for various applications such as remotesensing, medical imaging and weapon detection. In this work, the features of contourlet transform with Laplacian Pyramid and Hybrid Directional Filter Banks is combined with 2-D Discrete Wavelet Transform with an advantage of multi-scale localization, directionality and anisotropy. WeightedAverage method of Coefficient grouping is used for reconstructing the fused image with better structure character and edge information. The proposed method has considered MRI and CT images as inputs after image registration using Mutual Information Index of Information theory in terms of sizeand resolution information. The results produced by the proposed approach has delivered better fusion results than conventional wavelet, curvelet, contourlet, and some of the existing hybrid fusion methods.