Date of Award

5-2013

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Science

First Advisor

Quiming Zhu

Second Advisor

Zhengxin Chen

Third Advisor

Hamid Sharif

Abstract

In order to advance the field of computer vision in the direction of “strong AI”, it’s necessary to address the subproblems of creating a system that can “see” in a way comparable to a human or animal. Due to very recent advances in depth-sensing imaging technology, it is now possible to generate accurate and detailed depth maps that can be used for image segmentation, mapping, and other higher-level processing functions needed for these subproblems. Using this technology, I describe a method for identifying a moving object in video and segmenting the image of the object based on its motion. This creates a coarse vector field where each segment denotes a region of the object that is moving in the same general direction, rounded to the nearest 45 degrees. The approach described combines a conventional background subtraction algorithm, depth sensor data, and a biologically-inspired artificial neural circuit. In most cases the entire process can execute in near real time as the video is captured and is reasonably accurate.

Comments

A Thesis Presented to the Department of Computer Science and the Faculty of the Graduate College University of Nebraska In Partial Fulfillment of the Requirements for the Degree Master of Science University of Nebraska at Omaha. Copyright 2013 Corey A. Spitzer.

Download

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

COinS