Share Email Print
cover

Proceedings Paper

Multimodal sensor fusion for personnel detection
Author(s): Brian Maguire; Sachi Desai; Shafik A. Quoraishee
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Proposed herein is a framework between disparate multiple sensing modalities with correlated biometrics between the orthogonal modalities to aide in actor identification and discrimination within a particular defined scene. For the purposes of detection and classification, there are advantages and disadvantages to using either seismic or ultrasonic-sensing modalities individually to detect, identify, and classify successfully a subject to some degree. The seismic modality provides a clean, separable set of results for walkers close to the sensor but in general fails to be as successful at a larger stand-off due to ground wave attenuation and provides only some level of detection. The ultra-sonic modality provides stronger detection results from a greater stand-off distance in direct comparison to the seismic sensor due to higher fidelity of the signal. However, the resulting signature is harder to separate into individual walk-cycle events due to the broadband nature of the ultrasonic profile. In light of the advantages and disadvantages, it is desirable to utilize the capabilities of both sensors to create a framework for a robust singular solution for detection, identification, and classification of the actor(s) within the scene. In this particular case, data fusion techniques such as Hidden Markov Models and Dempster Shafer Theory provide a framework and methodology for exploiting each modality in an optimized and complementary fashion. In a probabilistic sense, this means that the chance of a successful detection is increased through a decision model that uses both the seismic and ultrasonic detection results to detect a target with high confidence.

Paper Details

Date Published: 24 May 2011
PDF: 11 pages
Proc. SPIE 8046, Unattended Ground, Sea, and Air Sensor Technologies and Applications XIII, 80460E (24 May 2011); doi: 10.1117/12.888874
Show Author Affiliations
Brian Maguire, U.S. Army Armament Research, Development and Engineering Ctr. (United States)
Sachi Desai, U.S. Army Armament Research, Development and Engineering Ctr. (United States)
Shafik A. Quoraishee, U.S. Army Armament Research, Development and Engineering Ctr. (United States)


Published in SPIE Proceedings Vol. 8046:
Unattended Ground, Sea, and Air Sensor Technologies and Applications XIII
Edward M. Carapezza, Editor(s)

© SPIE. Terms of Use
Back to Top