Improving the spatial resolution of hyperspectral image using panchromatic and multispectral images: An integrated method

Hyperspectral (HS) images have been extensively used in various fields; however, the low spatial resolution restricts its application. In this paper, an integrated HS image fusion method is proposed. An integrated model is established to express the relationships between the desired image and multi-source high-spatial-resolution observations. Maximum a posteriori (MAP) framework is employed to formulate the fusion model. On one hand, the proposed integrated HS fusion method can take full advantage of the more complementary spatial information from multi-source sensors to enhance the HS images. On the other hand, the complementary information of the high-spatial-resolution observations on spectral range is fully considered to maximize the spatial and spectral fidelity of all the HS bands. Furthermore, the proposed method is able to fuse the HS images for the challenging large-spatial-resolution-difference ratio. The HYDICE simulated and the ETM+ panchromatic (PAN), multispectral (MS), MODIS real datasets are used to verify the effectiveness of the proposed method.