Date of Award


Document Type


Degree Name

Master of Science (MS)

Legacy Department

Electrical Engineering

Committee Chair/Advisor

Birchfield, Stanley

Committee Member

Schalkoff , Robert

Committee Member

Hoover , Adam


This thesis examines the problem of person following. A person following algorithm can be separated into two distinct parts: the detection and tracking of a target and the actual following of a target. This thesis focuses mainly on the detection and tracking of a target person. For the purposes of this thesis a simple robot control architecture is used. The robot moves to follow the target in a straight line. No path planning is considered when executing robot movement.
This thesis aims to accomplish three tasks. First, the system should be able to track and follow a target when no occlusions occur. The non-occlusion scenarios should consider the target in environments with no other people, environments with other people present at different distances, and environments with other people present at similar distances. The second goal will be to track the target person through brief occlusions. The system should be able to detect when the target has been occluded, register the occlusion, and reacquire the target upon completion of the occlusion. The third and final goal of this thesis is to reacquire the target after a long term occlusion. The system must recognize that the target person has disappeared from the scene, wait for the target to reappear, and reacquire the target upon reappearance.
These goals will be accomplished using a generic person detector realized by a HOG person detector, a specific appearance model based on color histograms, a particle filter that will serve as an integrating structure for the tracker, and a simplistic robot control architecture.
In the following chapters I will discuss the motivation behind this work, previous research done in this area, the methods used in this thesis and the theory behind them. Experimental results will then be analyzed and discussion concerning the results and possible improvements to the system will be presented.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.