A Simple and Robust Destriping Algorithm for Imaging Spectrometers: Application to Modis Data

Content Provider	Semantic Scholar
Author	Bouali, Marouan
Copyright Year	2010
Abstract	Images provided by several imaging spectrometers are often contaminated with stripes. This artifact compromises the visual quality and radiometric integrity of measured data. Although a large number of destriping algorithms have been recently suggested, most of them provide results that display residual stripes if not strong distortion from the original signal. To overcome this issue, we introduce a robust methodology using a gradient-based iterative destriping algorithm (GBIDA). Statistical assumptions used in previous methods such as histogram matching are replaced with a more realistic geometrical consideration on the images spatial gradient. An iterative scheme is then used in order to isolate the striping effect from the original image prior to processing. Application of this method on MODIS level 1B images heavily affected with detector-to-detector stripes, mirror-side stripes and random stripes displays excellent qualitative and quantitative results; the visual quality of the images is greatly improved without
File Format	PDF HTM / HTML
Alternate Webpage(s)	http://www.asprs.org/wp-content/uploads/2013/08/Bouali.pdf
Alternate Webpage(s)	http://info.asprs.org/publications/proceedings/sandiego2010/sandiego10/Bouali.pdf
Language	English
Access Restriction	Open
Content Type	Text
Resource Type	Article