Image fusion using wavelet packet transform

In this paper, we especially focus on the question of handling multirescales images fusion based on pixels level Several approaches existing to image fusion are discussed respectively, of which the drawbacks are given. To overcome the faults we propose image fusion algorithm based on wavelet packet transform, which has advantages over the wavelet transform multiresolution analysis because of the decompositions procedure is applied to all the coarse scale approximations (low frequency) and detail information (high frequency). The inverse wavelet packet transform coefficients is calculated by the weighted factors, blurred edges information is eliminated and edges details is not damnified a bit thick simultaneous. For the high frequency sub-band the maximum coefficient values are gotten the strongest edges information is obtained. Finally, we compare this scheme with the others fusion approaches existing by simulation experiments. The results show that the approach using wavelet packet transform is satisfied subjectively and objectively.