Wavelet packet based fusion of IR and gray image

Image fusion is the method of combination of multiple images obtained from various image sensors so that we can have a better perception of a particular scene. In most of the applications image fusion is performed in Transform domain. Here a new algorithm is proposed where an image is converted into Approximate and Detail components by DWPT and their respective terminal nodes are obtained. Then the correlation- coefficient between two corresponding nodes is calculated and depending upon whether the correlation coefficient is positive, negative or zero, the terminal nodes are fused according to a fusion rule. After forming the fused nodes, the image is synthesized by using the haar Synthesis Filter Bank on the fused nodes. Results obtained are analyzed by calculating the performance parameters and there is no spatial distortion like DWT.