Labartkava_Gvarami.pdf (4.27 MB)
Download file

FPGA-Based Stereo-Vision System Mimicking Human Binocular Vision

Download (4.27 MB)
thesis
posted on 2021-05-22, 12:16 authored by Gvarami Labartkava

Human vision is a complex system which involves processing frames and retrieving information in a real-time with optimization of the memory, energy and computational resources usage. It can be widely utilized in many real-world applications from security systems to space missions. The research investigates fundamental principles of human vision and accordingly develops a FPGA-based video processing system with binocular vision, capable of high performance and real-time tracking of moving objects in 3D space. The undertaken research and implementation consist of: 1. Analysis of concepts and methods of human vision system; 2. Development stereo and peripheral vision prototype of a system-on-programmable chip (SoPC) for multi-object motion detection and tracking; 3. Verification, test run and analysis of the experimental results gained on the prototype and associated with the performance constraints; The implemented system proposes a platform for real-time applications which are limited in current approaches.

History

Language

eng

Degree

Master of Applied Science

Program

Electrical and Computer Engineering

Granting Institution

Ryerson University

LAC Thesis Type

Thesis