1 June 2006 Impulse noise removal with gradient adaptive neighborhoods
Author Affiliations +
Optical Engineering, 45(6), 067003 (2006). doi:10.1117/1.2213972
A new effective algorithm of impulse noise suppression is proposed. Conventional filtering schemes usually utilize a fixed shape for the moving window, such as a rectangle or circle. In contrast, the proposed algorithm exploits a signal-dependent shape for the moving window. We suggest a simple adaptive algorithm of impulse noise detection in monochrome images that takes into account the size of signal gradient neighborhoods and image statistics. Experimental results show superior performance of the proposed algorithm compared to that of conventional algorithms in terms of both subjective and objective criteria.
Mikhail G. Mozerov, Vitaly I. Kober, "Impulse noise removal with gradient adaptive neighborhoods," Optical Engineering 45(6), 067003 (1 June 2006). http://dx.doi.org/10.1117/1.2213972

Image filtering

Digital filtering

Gallium nitride

Image processing

Interference (communication)

Detection and tracking algorithms

Signal detection

Back to Top