Evaluation of spatial and spectral effectiveness of pixel-level fusion techniques

Abstract

Along with the launch of a number of very high-resolution satellites in the last decade, efforts have been made to increase the spatial resolution of the multispectral bands using the panchromatic information. Quality evaluation of pixel-fusion techniques is a fundamental issue to benchmark and to optimize different algorithms. In this letter, we present a thorough analysis of the spatial and spectral distortions produced by eight pan sharpening techniques. The study was conducted using real data from different types of land covers and also a synthetic image with different colors and spatial structures for comparison purposes. Several spectral and spatial quality indexes and visual information were considered in the analysis. Experimental results have shown that fusion methods cannot simultaneously incorporate the maximum spatial detail without degrading the spectral information. Atrous_IHS, Atrous_PCA, IHS, and eFIHS algorithms provide the best spatial-spectral tradeoff for wavelet-based and algebraic or component substitution methods. Finally, inconsistencies between some quality indicators were detected and analyzed.