Pleiades Satellite Remote Sensing Image Fusion Algorithm Based on Shearlet Transform

摘要

According to the characteristics of multi-spectral and panchromatic remote sensing images, this paper proposes a new image fusion algorithm based on nonsubsampled shearlet transform. Firstly, low frequency sub-band and high frequency directional sub-band is obtained by IHS color space conversion, and nonsubsampled shearlet transform apply to original multi-spectral images. Then, the adaptive weighted fusion rules is used to construct the low frequency sub-band coefficients, and regional consistency check and region-based local variance fusion rule are used to construct the high frequency sub-band coefficients. Lastly, the inverse of nonsubsampled shearlet transform and the inverse IHS transform are applied to the sub-band coefficients and then obtained the fused image. The comparison of simulation results with traditional fusion methods indicates that the proposed method have better fusion performance both in spectral characteristics and spatial resolution.