A directed search algorithm for setting the spectral-spatial quality trade-off of fused images by the wavelet à trous method

Abstract

This paper proposes a method to determine, in an objective and accurate way, the weighting factor (alfa) to be applied to the detailed panchromatic image information that will be integrated with the background multispectral image information to obtain the "best"; fused image with the same spatial and spectral quality. The fusion method is a weighting variant of the fusion algorithm based on the wavelet transform, calculated using the à trous (WAT) algorithm. The "alfa"; factor is determined, for each band of the multispectral source images using the simulated annealing (SA) search algorithm, which optimizes an objective function (OF) associated with both spatial and spectral quality measures for the fused images. The results obtained have demonstrated that for each one of the spectral bands there is an "alfa"; value that provides fused images with the optimal trade-off between the two qualities for any decomposition level value (n) of the wavelet transform.