Sub-pixel resolution thematic map based on multi-source data fusion

In this paper, super-resolution mapping method is used to enhance the spatial resolution of the original hyperspectral images (HSI) and process the mixed pixels, producing the sub-pixel resolution thematic map. Currently, super-resolution mapping usually selects only panchromatic images (PAN) or light detection and ranging images (LiDAR) to be fused with HSI, while this paper proposes a multi-source data fusion method for super-resolution mapping (MSDF), which first fuses HSI with PAN by pansharpening based on principal component analysis (PCA), followed by LiDAR by generalised fusion map method, and finally classifies the fused image by support vector machine (SVM) to generate the sub-pixel resolution thematic map. The proposed MSDF in this paper fuses three types of data sources (i.e., HSI, PAN and LiDAR) to obtain the fused image with full spatial-spectral-elevation information, which enhances the classification accuracy of the sub-pixel resolution thematic map.