Fusion of Multi-focus and Multi-exposure Colour Images Using Curvelet Transform Technique

Image fusion refers to the process of combining the relevant information from two or more images into a single highly informative image. The resulting fused image contains the salient information present in each of the input images. In this paper, an algorithm for fusing color images based on curvelet transform technique is being implemented. The other fusion techniques such as wavelet transform, Brovey, IHS, PCA have much less spatial information. This disadvantage is overcome by employing curvelet transform in the proposed work. In the literature discussed so far, only the monochrome image fusion using curvelet transform is considered. In this paper, colour images are fused using curvelet transform and the fused image preserves the vital colour information of the original images. In curvelet transform, the fused images have the same spectral resolution as the multispectral images and the same spatial resolution as the panchromatic image with minimum artifacts. It exhibit very high directional sensitivity, is highly anisotropic, represents edges better than wavelets, handles curve discontinuities well and is well suited for multi-scale edge enhancement. The different images such as multi-focused image, multi-exposure are fused into a new image to improve the information content. In the present fusion algorithm, the input registered colour images are fused using the curvelet transform. The fusion results are evaluated and compared according to four measures of performance - the Entropy (H), Root Mean Square Error (RMSE), Peak Signal to Noise Ratio (PSNR) and Correlation Coefficient (CC). These results are compared quantitatively with the wavelet transform technique.