Etudiant : JAAFANE SALAH EDDINE

Filière : LF Sciences Mathématiques et Informatique

Encadrant : Pr. SABRI ABDELOUAHED

Annèe : 2023

Résumé : We have created a Java application that uses deep learning techniques to detect banknotes . Additionally, we have developed two more models to provide multiple choices and preference to the user , these models have the potential later to be deployed and integrated into the app to enhance it’s functionality for visually disabled individuals : a U-Net segmentation model and a YOLOv7 object detection model . The U-Net model identifies the location of banknotes within an image, while the YOLOv7 model recognizes the denomination of the detected banknote . Our app can provide real-time assistance to visually disabled individuals , helping them to identify and count their banknotes quickly and easily , thus making a positive impact on the lives of those with visual impairments , improving their financial independence and overall quality of life .