Image fusion using a hybrid representation of empirical mode decomposition and contourlet transform

Abstract

This paper proposes an image fusion scheme based on a hybrid representation of empirical mode decomposition (EMD) and contourlet transform (CT), named the EMD-CT decomposition. The EMD-CT decomposition consists of three stages: the EMD stage, the Laplacian pyramid stage, and the multidirection analysis stage. As a result, the proposed EMD-CT shares high adaptivity of the EMD while owning multidirection analysis of the CT. For image fusion, fusion rules are applied on the EMD-CT representations of input images to produce a composite representation. The fused image is obtained by inversely transforming the composite EMD-CT representation. Experimental results show that the proposed fusion algorithm is more effective than fusion algorithms based on individual EMD or CT, and produces high fusion quality, especially for images with rich directional features such as edges and contours.