An efficient method for Blind Image Restoration using GAN
S.G.Gino Sophia, Syed Muhammad Umar, N. Visvas · 2022 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES) · 2022
The purpose of this project is to convert blind (or blurred) images to well-defined images using various stages of image processing such as preprocessing, image enhancement, segmentation and image restoration. The previous method involved Convolutional Neural Network (CNN) which was unable to identify noisy data appropriately using statistical methods. The main idea behind the project is to give a correct estimation of the image from a degraded version. In the proposed system incorporating Generative Adversarial Networks (GANs) to enhance the image thus, leading to a well-defined image for further study and development. There are some instances where the blind face (or blurred images) cannot be enhanced with human help so face image restoration is proposed to enhance such images in order to get clear images.