V. Aslantas [6] presented a new image fusion scheme using differential evolution algorithm. Images are divided into blocks. Optimal block size is selected using differential evolution algorithm. This algorithm is fast. DE involves 3 operations: initial population generation, mutation, crossover to choose optimum block size. Then focus measure of block calculated using spatial frequency, variance or sum-modified-laplacian measure. Higher value of focus measure, sharper is the image block. Thus sharpness value is compared of 2 corresponding blocks and shaper blocks are selected to construct a fused image. Then global sharpness value of imageis calculated. Larger the sharpness value better is the fused image. The process is repeated till a predefined condition is satisfied. For performance measure of fusion process, peaksignal to noise ratio (PSNR) and mutual information (MI) are used. Larger value of MI MSE and PSNR gives better fusion. Differential evolution (DE) algorithm is compared withgenetic algorithm (GE) and transform based methods like laplacian and wavelet. DE is reliable than GA based method. Fused image has higher quality than previous techniques.