The use of a computer to track objects has been a subject of interest for a few decades. The applications these algorithms may be applied to span a large number of fields; anything from homeland security to the study of animal behavior. In particular, visible tracking has been around the longest and has the largest library of algorithms available. Algorithms such as Mean Shift have become a standard for testing algorithms against. However, algorithms such as Mean Shift may work well for visible video data, infrared video data presents some issues beyond many visible algorithms. Infrared video gives certain advantages over visible, such as day/night tracking and camouflage detection. However, it also presents several issues as well. The detectors are more easily saturated, causing a temporary loss of data, as well as the drastic change in object appearance. These issues do not override the utility of infrared video being used for tracking purposes. This paper will go through some of the various applications of tracking, as well as the necessity of algorithm development in the infrared field. A couple algorithm metrics are also considered for a new basis for the testing of algorithms, as well as the introduction of a tracking algorithm testing platform.
Library of Congress Subject Headings
Remote sensing--Data processing; Infrared imaging--Data processing; Computer vision; Computer algorithms; Kalman filtering
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Ausfeld, Kyle, "Tracking of various targets in the infrared and issues encountered" (2012). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus