An Algorithm of Suppressing Image Noise Based on Wavelet Threshold Function and Improved Median Filtering

Abstract

For the presence of salt-and-pepper noise and Gaussian noise in image processing, an algorithm of suppressing image noise based on the wavelet threshold function and improved median filter is proposed. In this algorithm, salt-and-pepper noise is firstly estimated via the significant differences of pixel grayscale values between the salt-and-pepper noise and its surrounding, and then the 3×3 template is applied in order to filter out salt-and-pepper noise iteratively. Experimental results show that the peak signal to noise ratio (PSNR) is increased, the mean square error (MSE) is decreased and the image detail information is better preserved.