An independent wavelet reconstruction implementation for image fusion

A new wavelet reconstruction implementation for image fusion is proposed in this study. This reconstruction implementation is derived from Mallat´s algorithms. Using the reconstruction implementation, components for a panchromatic image can easily be modified through various image enhancement or filtering operations in order to improve the consistency with a multispectral image, prior to the fusion of images. In addition, more detailed components from the panchromatic image can be extracted and fused with the multispectral image thereby improving its spatial resolution. Visual and statistical analysis indicates that the new wavelet based image fusion method performs better at improving the spatial resolution while preserving the spectral characteristics of the multispectral imagery, when compared to conventional wavelet based image fusion methods.