One ability of the human visual system is the ability to identify and track moving objects. Examples of this can easily be seen in any sporting event. Humans are able to find an object in motion and track its current path and even predict a trajectory based on its current motion. Computer vision systems exist that are able to track an object in video, but usually these systems need to be instructed what the object to track is. As a way to further the work done by these computer vision systems, I present two additions to the work in the form of Adaptive Thresholding, a way to dynamically discover a threshold value of difference images, and a new method of blob tracking to further improve the accuracy of tracking blobs in video.
Library of Congress Subject Headings
Computer vision; Motion perception (Vision)--Computer simulation; Video surveillance; Video recordings--Data processing; Optical pattern recognition; Machine learning
Department, Program, or Center
Computer Science (GCCIS)
Case, Isaac, "Automatic object detection and tracking in video" (2010). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus
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 .C37 2010