A New Pansharpening Algorithm using Morphological Lifting Transform

Abstract

Pansharpening is a fusion technique that merges panchromatic (PAN) and multispectral (MS) images in order to enhance the spatial resolution of MS. Over the last two decades, a variety of approaches have been proposed either based on component substitution or multiresolution analysis. This paper proposes a new pansharpening algorithm using the morphological lifting transform (MLT) with region-based injection. The lifting structure with morphological operators provides a flexible tool to construct new morphological wavelet transforms. The morphological operators compared to the linear operators can extract the spatial details more effectively. In addition, an adaptive injection gain based on minimum mean squared error (MMSE) with k-means clustering is adopted. The proposed algorithm is validated in two datasets: Pleiades and Landsat 7. Experimental results show that the proposed method has advantages on enhancing the spatial details while retaining the spectral information. And it outperforms other state-of-the-art methods.