Panchromatic and Multispectral Image Fusion Combining GIHS, NSST, and PCA

Spatial and spectral information are essential sources of in remote sensing applications, the fusion panchromatic multispectral images effectively combines advantages both. Due to existence two main classes methods—component substitution (CS) multi-resolution analysis (MRA), which have different advantages—mixed approaches possible. This paper proposes a algorithm that generalized intensity–hue–saturation (GIHS) non-subsampled shearlet transform (NSST) with principal component (PCA) technology extract more spatial information. Therefore, compared traditional algorithms, this uses PCA transformation obtain structure components from PAN MS, can inject while maintaining high fidelity. First, is applied each band low-resolution (MS) (PAN) first calculate intensity MS. Then, image fused using NSST, used replace original component. Finally, obtained GIHS algorithm. Using urban, plants water, farmland, desert GeoEye-1, WorldView-4, GaoFen-7 (GF-7), Gaofen Multi-Mode (GFDM) as experimental data, method was tested evaluation mode references without five other classic algorithms. The results showed algorithms had better performances both preservation incorporation.