Hybrid Pan-Sharpening Method of Multispectral Satellite Image
Date Issued
2012
Date
2012
Author(s)
Chiu, Yen-Wei
Abstract
A typical multispectral sensor can offer both multispectral images and a single band panchromatic image. Panchromatic image has better spatial resolution but lack of spectral information; instead, multispectral images have red, green, blue even Near-Infrared band information with lower spatial resolution. In the application of remote sensing, for better classification result and visual presentation, we can “fusion” these two different images to acquire a high resolution multispectral image. We can use the fusion result to retrieve more information, or help us for further analysis or decision making.
Pan-sharpening is a specific research issue in image fusion. It focuses on fusing the panchromatic image and multispectral image from the identical sensor. The Pan-sharpening methods can divide into three categories: Pixel-based, Frequency-based, and hybrid method. This research will analysis some conventional pan-sharpening scheme, try to improve them from their advantages and disadvantages, and propose an improved hybrid pan-sharpening method. This research also proposes some spectral and spatial quantity assessment indices, to help fusion image user choose proper method for their application, or evaluate their result of image fusion.
Subjects
Multispectral image
Panchromatic image
Image fusion
Pan-sharpening
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-101-R99521110-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):8c63df9d1ac8250795e23492ab1c5657