A Weighted Image Fusion Approach Based on Multiple Wavelet Transformations

This paper, based on analyzing the characteristics and limitations of the classic pixel-level fusion algorithms, proposes a weighted image fusion approach based on multiple wavelet transformations. The approach firstly analyzes the negative impact aroused by the giant differences between luminance component of true colour image and panchromatic image, then, it adopts multiple wavelet transformations with different initial phase to resolve the negative impact. Experiments on Quick-bird images indicate that the approach proposed by this paper is feasible and get better performance than the classic pixel-level fusion algorithms with the reservation of spatial information and spectral information effectively.