Data fusion on remote sensing is one of important problems in current image processing. As the remote sensing data is massive, high-order redundant and non-negative, a new approach non-negative independent component analyses based on integer wavelet transform (IWT-ICA) is proposed and applied to remote image fusion between Landsat ETM+ panchromatic and CBERS multi-spectral images. After decomposing the remote sensing images by integer wavelet transform, the high frequency sub-band coefficients are done independent components analysis, and then the fused image is obtained by inverse integer wavelet transform. Visual and statistical analyses prove the new method not only significantly improves the fusion quality but also accelerates the operational speed compared to conventional principal component analysis and wavelet fusion techniques.
展开▼