Removal of Moiré Pattern Noise in Images using Median and Gaussian Filter

Moiré pattern noise in imaging introduces significant errors in qualitative and quantitative analysis of image. As the noise having complex origin; avoiding Moiré pattern noise is very difficult during the time of image data acquisition stage. Moiré pattern noise causes faint peaks in low frequency area, so using wavelet transform low frequency area is defined and threshold has been set manually. In this method we introduce a post processing technique for filtering Moiré pattern noise from images. This method includes a semiautomatic detection of the noise pixels by using a local median filter and the spectral noise peaks are eliminated using a Gaussian filter. The proposed MedianGaussian filtering framework shows good results for affected images and outperforms the existing method. Median-Gaussian filtering framework needs to set several parameters by user on a trial and error basis, and these parameters can be different for different types of image. Although the trial-and-error way works well for proving of principle, which is the aim of this project so appropriate selection of parameters is must. KeywordsMedian filter, Gaussian filter