|Acquisition Program: ||PEO Intelligence, Electronic Warfare and Sensors|
| ||The technology within this topic is restricted under the International Traffic in Arms Regulation (ITAR), which controls the export and import of defense-related material and services. Offerors must disclose any proposed use of foreign nationals, their country of origin, and what tasks each would accomplish in the statement of work in accordance with section 3.5.b.(7) of the solicitation.|| Objective: ||To develop algorithms that perform the detection and tracking of dismounts in a persistent, multiple-sensor urban surveillance scenario. The algorithms will integrate state-of-the-art advances in persistent, multiple-sensor surveillance and tracking, human activity modeling, pattern recognition techniques to recognize human activity, and automatic/assisted target detection and recognition. The system will be composed of distributed nodes and be able to track relevant targets through multiple sensors’ field of view.
|| Description: ||Over recent years, much research has taken place in the fields of dismount detection and tracking in terms of persistent surveillance in general with success limited to simple perimeter intrusion detection or variations from learned, regular pedestrian traffic patterns. Even more intractable is the problem of detecting single dismounts in urban environments, associative tracking of individual dismounts, and analysis to determine dismount activity in data from multiple sensors. Success is needed in associative tracking of dismounts in urban environments and determination of patterns of adversarial intent once effective tracking has been established. The innovation here over previous work is the ability to not only perform associative tracking of individual dismounts in urban environments (meaning many dismounts going about their normal routines) but also to track long enough to recognize adversarial intent. As of yet, no military system effectively maintains a long term, persistent, urban surveillance environment that can track and monitor the activities of dismounts through multiple sensors.
|| ||PHASE I: (Respondents are not required to develop hardware for program.) Will investigate, enhance, combine, and create algorithms and methodologies for persistent, multiple-sensor surveillance and recognition and modeling of human activity. Will provide specific and detailed testing plan focused on proving applicability. Will conduct limited tests.
|| ||PHASE II: Will conduct full interpretation/classifier system tests. Will demonstrate functioning and utilizable prototype system. System will perform successful persistent, multiple-sensor surveillance and enable assisted human activity modeling and interpretation.
|| ||PHASE III: Commercialization of technology would involve all types of surveillance--end of state would be a viable system for interpretation of human intent. A specific military application of this would be base protection by the Army in Afghanistan and Iraq. Most likely customers of technology would be commands seeking to utilize surveillance capabilities. Phase 3 funding should be provided by these organizations. In addition, results of Phase 2 will be utilized in NVESD persistent surveillance efforts. Also, upon success of Phase 2 results, PM RUS has expressed interest in possible Phase 3 utilzation in their Persistent Surveillance Threat Detection System (PTDS). There is currently one system in Theater and six more on contract.Transition will be achieved by direct implementation of Phase 2 software package in existing sensor hardware. Commercial applications would include building and property security--the remote supervising of the perimeter of an office or physical plant for hostile action, or the remote supervising of crowds in shopping malls, sports stadiums, or casinos for illegal activity.
|| References: ||1) Qing Cao, Tian He, and Tarek F. AbdelZaher. uCast: Unified Connectionless Multicast for Energy Efficient Content Distribution in Sensor Networks. IEEE Transactions on Parallel and Distributed Systems, 2006.
2) V. Calloway, R. Hodges, S. Harman, A. Hume, D. Beale:
Vehicle Tracking using a Network of Small Acoustic Arrays.
Proceedings of SPIE—v.5429, Signal Processing, Sensor Fusion, and Target Recognition XIII, 2004.
3) Alexander Kott, Michael Ownby: Adversarial reasoning: challenges and approaches, Proceedings of SPIE – v.5805, Enabling Technologies for Simulation Science IX, 2005.
4) Dongxu Li, Jose B. Cruz, Jr., Genshe Chen, Chiman Kwan, and Mou-Hsiung Chang: A Hierarchical Approach to Multi-Player Pursuit-Evasion Differential Games, Proceedings of the 44th IEEE Conference on Decision and Control, 2005.
5) Sanjeev Arulampalam, Simon Maskell, Neil Gordon, Tim Clapp: A Tutorial on Particle Filters for On-line Non-linear/Non-Gaussian Bayesian Tracking.
IEEE Transactions on Signal Processing, v.50, #2, 2002.
|Keywords: ||dismounts, human tracking, surveillance, multi-sensor fusion, human activity modeling, artificial intelligence, urban warfare, automatic target recognition|