10 February 1999 Statistical sensor fusion analysis of near-IR polarimetric and thermal imagery for the detection of minelike targets
Author Affiliations +
Proceedings Volume 3534, Environmental Monitoring and Remediation Technologies; (1999) https://doi.org/10.1117/12.339013
Event: Photonics East (ISAM, VVDC, IEMB), 1998, Boston, MA, United States
Abstract
We present an analysis of statistical model based data-level fusion for near-IR polarimetric and thermal data, particularly for the detection of mines and mine-like targets. Typical detection-level data fusion methods, approaches that fuse detections from individual sensors rather than fusing at the level of the raw data, do not account rationally for the relative reliability of different sensors, nor the redundancy often inherent in multiple sensors. Representative examples of such detection-level techniques include logical AND/OR operations on detections from individual sensors and majority vote methods. In this work, we exploit a statistical data model for the detection of mines and mine-like targets to compare and fuse multiple sensor channels. Our purpose is to quantify the amount of knowledge that each polarimetric or thermal channel supplies to the detection process. With this information, we can make reasonable decisions about the usefulness of each channel. We can use this information to improve the detection process, or we can use it to reduce the number of required channels.
© (1999) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Robert A. Weisenseel, William Clement Karl, David A. Castanon, Charles A. DiMarzio, "Statistical sensor fusion analysis of near-IR polarimetric and thermal imagery for the detection of minelike targets", Proc. SPIE 3534, Environmental Monitoring and Remediation Technologies, (10 February 1999); doi: 10.1117/12.339013; https://doi.org/10.1117/12.339013
PROCEEDINGS
9 PAGES


SHARE
Back to Top