A Novel approach to denoise an image using CNN
R. Nagendra, D. Harika, Kallivettu Ganesh, H. Hemanth, T. Santhosh, U. Venkat Munireddy · 2023
Due to a variety of environmental and human factors, real-world photographs frequently contain noise. De-noising is a method for taking out noise from an image. De-noising images faces significant challenges because of the sources of noise. Gaussian, impulse, salt-and-pepper, and speckle noise are some examples of complicated sources of noise in imaging. In the realm of visual de-noising, Convolutional Neural Networks (CNN) have attracted an increasing amount of interest. Several CNN approaches can be used to denoise images. For the evaluation of these approaches, various datasets were used. The various CNN algorithms used for image de-noising are studied in this study. Many CNN architectures have been used over the years to denoise digital images. RIDNet and Autoencoder, two cutting-edge CNN architectures, are used here.