1. Impulse noise reduction based on improved median ϵ-filter.
- Author
-
Shaneh, Mahdi and Mahyari, Arash Golibagh
- Subjects
- *
ELECTRONIC noise , *IMAGE processing , *PERFORMANCE evaluation , *MEDIAN filters (Electronics) , *ACQUISITION of data , *NONLINEAR theories - Abstract
Impulse noise reduction has been a challenge for Image Processing researchers for many decades. Median filter had been the most popular denoising methods. However, owing to its edge jittering and removing image detail properties, numerous efforts have been taken in order to improve its performance. Median ϵ-filter is one of the modified median filters. Median ϵ-filter and median filter behave in a like manner in the presence of noise. Thus, it has all disadvantageous things of median filter. In this article, we improve median ϵ-filter's efficiency by giving weights to the outcomes of ϵ-filter. The experimental results show the excellent performance of our method. Other than that, in the absence of noise, our method preserves the original image's details more than others. [ABSTRACT FROM PUBLISHER]
- Published
- 2012
- Full Text
- View/download PDF