Fast Rank Algorithms with Multiscale Histograms Lazy Updating
Publication type
Conference paper
Author(s)
M. Storozhilova, D. Yurin
Publication date
November 2011
Conference
8th Open German-Russian Workshop “Pattern Recognition and Image Understanding” (OGRW-8-2011)
Page(s)
280-283
Address
Lobachevsky State University of Nizhny Novgorod, November 21-26, 2011
Abstract
Rank algorithms propose good solutions for image smoothing and impulse noise removing. In our previous work on the subject we have proposed methods for fast computations of εV and KNV neighborhood average based on multiscale histograms. In this paper a new method for multiscale histograms fast updating is described and analyzed. Performance comparisons of the combined methods with fastest known median filtering algorithm are given. We have achieved the processing speed for εV and KNV neighborhood average algorithms only a few times slowly than the fastest known median filtering algorithm.
