Abstract:

To overcome the poor directional selectivity of wavelet transform, a new self-adaptive multi-sensor image fusion algorithm was proposed based on the dual-tree complex wavelet-Contourlet transform.Using this algorithm,the panchromatic and multispectral images were first decomposed and transformed,and then different fusion rules were used according to different frequency characteristics: the self-adaptive regional energy coefficients added fusion rule were used for the lowfrequency and the regional feature self-adaptive fusion rule for the high-coefficients.The final fusion images was obtained by reconstruction.Compared with other fusion algorithms, the results showed that the self-adaptive image fusion algorithm based on the dual-tree complex wavelet-contourlet transform was feasible and effective.

Key words:  multisensor images, dual-tree complex wavelet transform, Contourlet transform, self-adpative, fusion algorithm