Comparative Study of High Density Salt and Pepper Noise Removal (Spatial Domain Methods used in Image Processing)

This paper deals with the study of Salt& Pepper noise and different de-noising methods to remove them. Noise densities varies from 10 to 98 % in an image. Various filters such as Mean Filters, Order-Statistic Filters and Adaptive filter are used for image de-noising. From Mean Filter Arithmetic Mean Filter, Max and Min Filter from Order-Statistic Filters and Adaptive mean Filter from Adaptive Filters and Median filters respectively. Comparison of different noise removal technique is based on three parameters Mean Absolute Error (MAE), Mean Square Error (MSE), and Peak Signal Noise Ratio (PSNR). The mean absolute error is a quantity used to measure how close forecast or predictions are to the eventual outcomes. The mean square error of an estimator is one of many ways to quantify the difference between an estimator and the true value of quantity being estimated. MSE is a risk function, corresponding to the excepted value of the squared error loss or quadratic loss. MSE measures the average of the square of the error. The error is the amount by which the estimator differs from the quantity to be estimated. The phrase Peak Signal to Noise Ratio, often known as PSNR, is an engineering term for the ratio between maximum possible power of a signal and the corrupting noise that affects the fidelity of its representation. Because many signals have a very wide dynamic range, PSNR is usually expressed in terms of the logarithmic decibel scale. Our quantitative results with help of MATLAB simulations show that for lower noise densities to medium-high densities (10 to 98 %), Max filter gives the optimum performance followed by adaptive filter, Median filters. For noise densities from 60 % to 80 %, Min Filter showed best results, followed by Max and Adaptive median filter respectively. For higher noise values (greater than90 %) the Min Filter showed the best results followed by Adaptive mean filter and median filter.