Research output: Contribution to conference › Paper
83Downloads
(Pure)
Abstract
Aperture filters are a recently introduced class of nonlinear filters used in image processing. In this paper we present a new approach for aperture filter design, improving operator performance with respect to the MSE measure by releasing some of the operator constraints without losing statistical estimation accuracy. With the use of the proposed methods an average of 34% MSE reduction was achieved for deblurring, whereas a standard aperture operator reduced the error by only 10% on the average.
abstract = "Aperture filters are a recently introduced class of nonlinear filters used in image processing. In this paper we present a new approach for aperture filter design, improving operator performance with respect to the MSE measure by releasing some of the operator constraints without losing statistical estimation accuracy. With the use of the proposed methods an average of 34{\%} MSE reduction was achieved fordeblurring, whereas a standard aperture operator reduced the error by only 10{\%} on the average.",
keywords = "filter constraints , apertures",
author = "S. Marshall",
year = "2007",
language = "English",
pages = "1492--1496",
note = "15th European Signal Processing Conference, EUSIPCO 2007 ; Conference date: 03-09-2007 Through 07-09-2007",
2007. 1492-1496 Paper presented at 15th European Signal Processing Conference, Poznan , Poland.
Research output: Contribution to conference › Paper
TY - CONF
T1 - Releasing aperture filter constraints
AU - Marshall, S.
PY - 2007
Y1 - 2007
N2 - Aperture filters are a recently introduced class of nonlinear filters used in image processing. In this paper we present a new approach for aperture filter design, improving operator performance with respect to the MSE measure by releasing some of the operator constraints without losing statistical estimation accuracy. With the use of the proposed methods an average of 34% MSE reduction was achieved fordeblurring, whereas a standard aperture operator reduced the error by only 10% on the average.
AB - Aperture filters are a recently introduced class of nonlinear filters used in image processing. In this paper we present a new approach for aperture filter design, improving operator performance with respect to the MSE measure by releasing some of the operator constraints without losing statistical estimation accuracy. With the use of the proposed methods an average of 34% MSE reduction was achieved fordeblurring, whereas a standard aperture operator reduced the error by only 10% on the average.
KW - filter constraints
KW - apertures
UR - http://www.eurasip.org/Proceedings/Eusipco/Eusipco2007/Papers/c3l-b03.pdf