Electronic Theses and Dissertations

Identifier

2594

Shahinur Alam

Date

2016

Document Type

Thesis

Degree Name

Master of Science

Major

Electrical and Computer Engr

Concentration

Computer Engineering

Committee Member

Mohammed Yeasin

Committee Member

Lavonnie Perry Claybon

Committee Member

Peter S Lau

Abstract

A system capable of detection and localization of objects of interest in a semi-structured environment will enhance the quality of life of people who are blind or visually impaired. Towards building such a system, this thesis presents a personalized real-time system called O'Map that finds misplaced/moved personal items and localizes them with respect to known landmarks. First, we adopted a participatory design approach to identify users’ need and functionalities of the system. Second, we used the concept from system thinking and design thinking to develop a real-time object recognition engine that was optimized to run on low form factor devices. The object recognition engine finds robust correspondences between the query image and item templates using K-D tree of invariant feature descriptor with two nearest neighbors and ratio test. Quantitative evaluation demonstrates that O'Map identifies object of interest with an average F-measure of 0.9650.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.

Download

Share

COinS