| | |
Summary: IEEE SIGNAL PROCESSING LETTERS, VOL. 11, NO. 3, MARCH 2004 363
Effective Impulse Detector based on
Rank-Order Criteria
Igor Aizenberg, Member IEEE, and Constantine Butakoff
Abstract-- A new impulsive noise detection technique is pre-
sented here. Preserving edges and details in the process of impul-
sive noise filtering is an important problem. To avoid image
smoothing, only corrupted pixels must be filtered. In order to
identify the corrupted pixels, a new impulse detector is proposed.
This detector is based on a comparison of signal samples within a
narrow rank window by both rank and absolute value. It is effi-
cient, very fast, and can be used with any filter (even iteratively),
without smoothing an image.
Index Terms--impulse detection, impulsive noise, median fil-
ter.
I. INTRODUCTION
mpulsive noise filtering is an important field in image
processing. As a rule, impulsive noise can significantly
damage an image. A commonly used way to filter impul-
sive noise is median filtering [1]. Usually it is possible to
|
