Classification-based fusion of IKONOS 1-m high-resolution panchromatic image and 4-m multi-spectral images

This paper presents a new classification-based fusion (CL-based fusion) scheme, which is used throughout this paper as a synonym of the "pixel-level fusion", according to the terminology [1]. It can be applied to merge low-resolution images and co-registered high-resolution image. The key of this trial is as follows: (1) Geometrical co-registration IKONOS 1-m high-resolution panchromatic image and 4-m multi-spectral images. (2) Classification of the high-resolution image together with the low-resolution images using supervised classification algorithm. (3) According to each class histogram of high-resolution image, adjustment of the corresponding spectral values of single multi-spectral image. (4) HIS transform of the adjusted multi-spectral images using cylindrical coordinates and acquirement of the fusion images. The proposed method can merge the class with high-resolution image and multi-spectral images, accounting for the importance of prior knowledge [2], and try to add classification as prior knowledge to fusion. In this trial, the selected Peking image of IKONOS will be used, which will bring us convenience when verifying our algorithm.