A novel multimodal image fusion method using Shift invariant Discrete Wavelet Transform and Support Vector Machines

In this paper, a multimodal image fusion technique using Shift invariant Discrete Wavelet Transform (SIDWT) and Support Vector Machines (SVM) suitable for surveillance applications is proposed. This technique uses SIDWT for multiresolution decomposition as it is translation invariant. A Support Vector Machine is trained to select the coefficient blocks with significant features, extracted from the SIDWT coefficients. The corresponding selected coefficients are used in forming the composite fused coefficient representation. The proposed method is tested for a number of multimodal images and found to outperform other traditional image fusion algorithms in terms of the various fusion metrics. Experimental results show that the quality of the fused image is significantly improved for multimodal images.