Image pansharpening: Comparison of methods based on multiresolution analysis and component substitution

In this paper, we propose to compare pansharpening techniques based on multiresolution analysis (MRA) and component substitution (CS). The appropriate way is to show the difference on methods which are identical in how to incorporate spatial details into multispectral (MS) data. Our study relies on the Gram-Schmidt (GS) transform technique that is an orthogonal decomposition. Thus, we focus on the two modalities of GS transform (GS1 and GS2), in order to point out the difference between MRA and CS techniques. Thus, we estimate quantitative indexes carried out on fused MS results. Eventually, we present and discuss the fusion products on Pleiades data set.