Human activity discrimination for maritime application

Evelyn Boettcher; Dawne M. Deaver; Keith Krapels

doi:10.1117/12.776291

Translator Disclaimer

11 April 2008 Human activity discrimination for maritime application

Evelyn Boettcher, Dawne M. Deaver, Keith Krapels

Author Affiliations +

Proceedings Volume 6941, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XIX; 69410G (2008) https://doi.org/10.1117/12.776291
Event: SPIE Defense and Security Symposium, 2008, Orlando, Florida, United States

Abstract

The US Army RDECOM CERDEC Night Vision and Electronic Sensors Directorate (NVESD) is investigating how motion affects the target acquisition model (NVThermIP) sensor performance estimates. This paper looks specifically at estimating sensor performance for the task of discriminating human activities on watercraft, and was sponsored by the Office of Naval Research (ONR). Traditionally, sensor models were calibrated using still images. While that approach is sufficient for static targets, video allows one to use motion cues to aid in discerning the type of human activity more quickly and accurately. This, in turn, will affect estimated sensor performance and these effects are measured in order to calibrate current target acquisition models for this task. The study employed an eleven alternative forced choice (11AFC) human perception experiment to measure the task difficulty of discriminating unique human activities on watercrafts. A mid-wave infrared camera was used to collect video at night. A description of the construction of this experiment is given, including: the data collection, image processing, perception testing and how contrast was defined for video. These results are applicable to evaluate sensor field performance for Anti-Terrorism and Force Protection (AT/FP) tasks for the U.S. Navy.

Citation Download Citation

Evelyn Boettcher, Dawne M. Deaver, and Keith Krapels "Human activity discrimination for maritime application", Proc. SPIE 6941, Infrared Imaging Systems: Design, Analysis, Modeling, and Testing XIX, 69410G (11 April 2008); https://doi.org/10.1117/12.776291

ACCESS THE FULL ARTICLE