Douzi Youssef, Benabdellah Mohammed, Azizi Abdelmalek, Hajji Tarik, Jaara El Miloud

Zoom and Restoring of Digital Images with Artificial Neural Networks

We will show in this work, the use of artificial neural networks (ANN) as a new way of doing zoom operations on digital images. The objective is to make the presentation and description of four new approaches to zoom based on the use of ANN. These approaches will allow us to achieve regional and global zoom on a digital image. We will use a supervised learning algorithm for our ANN. We will use a database of learning formed by a set of digital images and their versions zoomed carefully by software zoom. We will also even a brief on ANN, their working principle and the famous zoom digital images methods. In this work, we will also prove the effectiveness of ANN as a good solution to the problem of restoring digital images defective due to zooming.

Accéder au lien