Author

Raja Jain

Abstract

In this work, moving foreground objects are extracted from video. The video camera can be either stationary or panning. The motion information extracted from consecutive frames of video is used for object segmentation, and an appearance model of the moving object is created. No assumptions are made concerning the object model, resulting in segmented objects of varying shapes and sizes. Finally, relationships between objects in the image are evaluated toward the goal of higher level understanding of videos.

Library of Congress Subject Headings

Computer vision; Video recordings--Data processing; Image analysis; Image processing; Pattern recognition systems

Publication Date

2006

Document Type

Thesis

Department, Program, or Center

Computer Science (GCCIS)

Advisor

Gaborski, Roger

Advisor/Committee Member

Reynolds, Carl

Advisor/Committee Member

Anderson, Peter

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: TA1634 .J34 2006

Campus

RIT – Main Campus

Share

COinS