Simplified Filtering Technique to Edge Reconstruction and Comparison of Original Image in Low to High Density Noise Values

From last twenty years, removing noise from the image without any losses like over smoothed, blurred or spoiling edges and lines is challenging problem for researchers. When transmitting images through the network cables or channels the image may contaminate because of the storage characteristics. Even a small intensity values may change the entire view of an image. The change of intensity values is called Noise. There are different noises like Salt& Pepper Noise, Gaussian Noise, Speckle Noise, and White Noise. The pre-processing technique approach is of two types, they are filtering approach and wavelet based approach. There are different existing filtering techniques are there like Median Filter, Standard Median Filter, Adaptive Median Filter and Decision Based Algorithm. These traditional filtering techniques are proved to the best at Low noise densities. In this paper we present a filtering technique which will work efficiently at low noise densities as well as high noise densities. This filtering technique will perform in two states, firstly, eliminating the corrupted pixels based on centre waited pixel, finally reconstruction the original image without disturbing the original image edges. Simulation results are quantitatively measured as PSNR (Peak signal to noise ratio), MSE (Mean Square Error) and MAE (Mean Absolute Values).