08-01-2014, 12:55 PM
Salt-and-Pepper Noise Removal from Image by Median type Noise Detector
Abstract
Noise is the main drawback of transmission and reception. In the transmission noise are added due to system and atmosphere .So the quality of picture is decrease before applying image analysis processes. Noise should be removed while keeping the fine detail of the image intact.This paper proposes a two-dimensionalmethod for removing salt-and-pepper impulse noise. In the first phase, an adaptive median filter is used to identify pixels which are likely to be contaminated by noise(noise candidates). In the second phase, the image is restored using a specialized regularization methodthat applies only to those selected noise candidates. In terms of edge preservation and noise suppression,our restored images show a significant improvement compared to those restored by using just nonlinearfilters or regularization methods only. Our scheme can remove salt-and-pepper-noise with noise level as high as 90%
Abstract
Noise is the main drawback of transmission and reception. In the transmission noise are added due to system and atmosphere .So the quality of picture is decrease before applying image analysis processes. Noise should be removed while keeping the fine detail of the image intact.This paper proposes a two-dimensionalmethod for removing salt-and-pepper impulse noise. In the first phase, an adaptive median filter is used to identify pixels which are likely to be contaminated by noise(noise candidates). In the second phase, the image is restored using a specialized regularization methodthat applies only to those selected noise candidates. In terms of edge preservation and noise suppression,our restored images show a significant improvement compared to those restored by using just nonlinearfilters or regularization methods only. Our scheme can remove salt-and-pepper-noise with noise level as high as 90%