Document Details

Document Type : Thesis 
Document Title :
A Deep Learning-based Obstacle Classification Method from Partial Visual Information: Application to the Assistance of Visually Impaired People
طريقة للتصنيف من العلامات المرئية الجزئية عبر التعلم العميق: تطبيق لمساعدة ضعاف البصر
 
Subject : Faculty of Computing and Information Technology 
Document Language : Arabic 
Abstract : Hidden text: The abstract may be included at the discretion of the supervisor. Among the challenges that face visually impaired people is how to navigate safely, recognize encountered obstacles, and move independently from one location to another in unknown environments. Obstacle detection and classification are among the most challenging difficulties that hinder a visually impaired person from performing daily tasks. By proposing a solution towards overcoming these challenges, this work will be of most importance to visually impaired people. In this work, we propose a consistent, reliable and robust smartphone-based method to classify obstacles in unknown environments from partial visual information based on Multi-layer Perceptron (MLP) technique. Our proposed method deals with high levels of noise and bad resolution in the captured frames. In addition, it offers a maximum flexibility to the user and use the least expensive equipment possible. Moreover, our method, leveraging on deep-learning techniques, enables to semantically categorize the detected obstacles in order to increase the awareness of the explored environment. The efficiency of the work is measured by many experiments studies on different complex and unknown scenes. It records high accuracy of [90.2%]. 
Supervisor : Dr. Salma Mohamad Kammoun 
Thesis Type : Master Thesis 
Publishing Year : 1440 AH
2019 AD 
Co-Supervisor : Dr. Manar Sayed Salama 
Added Date : Tuesday, January 22, 2019 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
وفاء سعد الشهريAlshehri, Wafa SaadResearcherMaster 

Files

File NameTypeDescription
 43920.pdf pdf 

Back To Researches Page