Multi-directional decomposition and directional contrast based remote sensing images fusion

Abstract

Multi-directional decomposition and directional contrast based remote sensing images fusion algorithm is presented in this paper. Based on the properties of band contrast processing and contrast orientation of visual receptive fields, multi-directional contrast is introduced. The fusion procedure is as follows: first, Nonsubsampled Contourlet Transform of Panchromatic and multi-spectral images, second, construction of directional contrast sequences at different decomposition levels and different directions, third, construction of fused Nonsubsampled Contourlet Transform coefficients with multi-directional contrast fusion rule. Finally, a fused high spatial resolution multi-spectral image can be obtained with inverse Nonsubsampled Contourlet Transform. Experiment results show that multi-directional decomposition and directional contrast based remote sensing images fusion method can provide a very good fusion image. This method is better than ordinary wavelet transform based directional contrast fusion method. The fused multi-spectral image can provides more spatial information and preserve more spectral information. © 2008 SPIE.