| Summary: | 碩士 === 國立成功大學 === 電腦與通信工程研究所 === 104 === Pansharpening is the technique that fuses high spectral resolution multispectral (MS) images and a high spatial resolution panchromatic (PAN) image to create a high spatial and high spectral resolution fused image. In this Thesis, a two-phase pansharpening algorithm is proposed. First, the spatial resolution of MS images is up-scaled by super-resolution technique with the relationship between the PAN and MS images. Second, the algorithm based on the Intensity-Hue-Saturation method, and the intensity component is estimated by multiple linear regression between the PAN and MS images. The intensity component is replaced by the PAN image according to edge strength. Thus, the fused result can avoid injecting excessive details and noise in the flat area of land surface, such as pools and roofs. The experiment results show that the proposed algorithm achieves better visual quality and less artifact.
|
|---|
