Title: Vector median filter based on non-causal linear prediction for detection of impulse noise from images

Authors: Khumanthem Manglem Singh

Addresses: Department of Computer Science and Engineering, National Institute of Technology Manipur, Imphal – 795001, India

Abstract: A new switching vector median filter based on the non-causal linear prediction for the detection of impulse noise from colour images is presented. The proposed filter utilises non-causal linear prediction coefficients obtained from the block-by-block autocorrelation function to find prediction error by taking difference between the vector pixel value at the centre of the filter window and predicted vector pixel value at that location. The vector pixel at the centre of the filter window is replaced by vector median if the error is bigger than a pre-defined threshold. Extensive simulation experiments indicate that the new vector median filter outperforms the vector median filter as well as other techniques currently used to eliminate impulse noise from colour images.

Keywords: vector median filter; impulse noise; non-causal linear prediction; autocorrelation function; block-by-block; constrained intrachannel predictor; signal-dependent mean filter; rank-ordered mean filter; rank-conditioned VMF; threshold VMF; RCTVMF; centre-weighted VMF; CWVMF; switching VMF; colour images; image processing.

DOI: 10.1504/IJCSE.2012.049729

International Journal of Computational Science and Engineering, 2012 Vol.7 No.4, pp.345 - 355

Received: 24 Dec 2010
Accepted: 19 May 2011

Published online: 22 Sep 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article