Impulse noise is a spark that affects the contents of digital images. The proposed method detects the random valued impulse noise efficiently. Then the detected noisy pixels are restored by the median of neighbouring noise-free pixels. In a detection window of this proposed method, four directions are considered. From that, optimal direction is obtained by standard deviation. Threshold value is calculated by finding normalised distance between original pixel and other pixels in the optimal direction. The threshold value is used as a measure to detect whether the tested pixel is noisy or noise-free pixel. More edge pixels can be detected if the accurate or optimal direction of the edge is determined. The noisy pixel that has small deviations with the pixels in the optimal direction is seems like the original pixel. Here, in detection and in filtering, the window size is adaptive which depends on noise density in the detection window frame. The optimum threshold limit is fixed as 0. 8 by using normalised distance between the central pixel and pixels in the optimum direction in first iteration. In second iteration, the threshold value should be kept very close to the value zero to remove the undetected noisy pixel and that was found as 0. 4. In second iteration the noisy pixels will be mostly eliminated. It is found that the proposed method gives better results when compared to adaptive median filter, progressive switching median filter, in terms of PSNR and MSE values and output images are compared using MATLAB.

Optimal direction PSNR MSE MATLAB