Wavelet Based Technique For Removal Of Multiple Noises Simultaneously
Abstract: Denoising is important pre-processing tasks for various image processing. Image noise is the random variation of
brightness and color information in images produced by the scanner and digital camera. This paper presents a novel
approach for simultaneously removing the speckle and Salt-n-pepper noise from a single image by using the median filter.
This paper proposes an adaptive, data-driven threshold for image denoising via wavelet soft thresholding. The threshold is
derived in a Bayesian framework and also using the MAD (mean absolute difference) value of fast multidirectional filter
bank which improves the radial frequency resolution of the image by addition decomposition in the high frequency band.
Denoising performance of median filter will be compared with discrete wavelet transform and wiener filter. The
performance of median filter using parameter metrics PSNR (Peak Signal to Noise Ratio) and coefficient of correlation
(Coc) is also analyzed.