Comparision of Image Denoising using Convolutional Neural Network (CNN) with Traditional Method

Image Denoising plays a vital role in various applications in the present scenario like image segmentation, classification, restoration, etc., Image denoising will remove noise from the image and restore the image with high quality. Image has interrupted by noise through different ways like extrinsic (i.e., environment) or intrinsic (i.e., like sensors) conditions. There are different algorithms for image denoising. The proposed research work utilizes quantitative analysis, i.e., PSNR metric is considered for further comparison. In our proposed method, comparison of image denoising using convolutional neural network (CNN) model with general traditional method like wavelet based model. The analysis is done by adding Gaussian noise with different variance and calculate Peak Signal to Noise Ratio (PSNR) and Signal to Noise Ratio (SNR) for each images using convolutional neural network (CNN) and compared it with traditional wavelet based model. After the analysis, the final results show that the convolutional neural network (CNN) model gives better results when compared to wavelet based model. CNN method have higher PSNR and SNR value, when compared to wavelet based model.