A New Evaluation Protocol for Image Pan-sharpening Methods

Overview & Synopsis

Pan-sharpening consists in fusing the spatial and spectral characteristics of panchromatic and multispectral (MS) images to get synthesized MS images. When such a fusion technique is proposed, it is delicate and important to evaluate its results. Generally, to evaluate the pan-sharpening methods both spectrally and spatially, a variety of quality indexes are available. Although, spectral indexes play a more important role than spatial ones to assess the fusion methods, spatial quality is important too. In this paper, a new protocol is proposed to evaluate pan-sharpening methods. This evaluation, by considering both spectral and spatial indexes, facilitates, reduces and even avoids any visual analyses, and allows automatic classification when comparing fusion methods.