An Efficient Image Denoising Technique for Unprocessed Raw Images using Combine Linear and Non-Linear Filtering

Abstract

Image denoising is used to improve the accuracy and quality of an image. Removing noise from the original image is still challenging for researchers. In this research, an efficient algorithm capable of removing noise from "unprocessed" or raw images is proposed. The algorithm supplants the noise by the median of averages found from a special combination of the pixels. After that, to evaluate the performance of image authors has calculated the Signal to Noise Ratio (SNR), the Mean Square Error(MSE), Root Mean Square Error (RMSE), the Root Mean Square Signal to Noise Ratio (RMS SNR), Image Fidelity (IFY). Finally, the proposed filtering technique gives a better result with comparison to other existing filtering techniques (Median, Average, Mean, etc).