You are here

A RSTA Algorithmic Framework with Uncertainty Management

Award Information
Agency: Department of Defense
Branch: Army
Contract: N/A
Agency Tracking Number: 36717
Amount: $98,087.00
Phase: Phase I
Program: SBIR
Solicitation Topic Code: N/A
Solicitation Number: N/A
Solicitation Year: N/A
Award Year: 1997
Award Start Date (Proposal Award Date): N/A
Award End Date (Contract End Date): N/A
Small Business Information
5506 Shipley Ct
Centreville, VA 20120
United States
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Dr. Qinfen Zheng
 (703) 803-8841
Business Contact
Phone: () -
Research Institution

One of the most challenging information management problems related to the Reconnaissance, Surveillance and Target Acquisition (RSTA) missions is the real-time spatio-temporal fusion of target-specific cues from a netwotk of multiple sensor platforms. A robust algorithmic framework is necessary to involve the multiple agents together and achieve the optimal trade-off between low false alarm rate and high detection and recognition rate. As a technology transfer effort supported by Professor Rama Chellappa at the University of Maryland, Cyber Solutions Inc. proposes in this Phase I effort to develop an end-to-end algorithmic framework involving a suite of RSTA algorithms developed at the University of Maryland, College Park, MD. This includes an image stabilization technique (Zheng and Chellappa, 1993), a tracking algorithm (Zheng and Chellappa, 1995), and a multi-sensor registration technique (Zheng and Chellappa, 1992). We will develop a simple target recognition algorithm using a view-dependent model based neural net for Phase I. Unconstrained Automatic Target Recognition (ATR) is a complex task by itself. For sensor fusion, we plan to choose an uncertainty reasoning framework: a framework to integrate target hypotheses from the multiple sensor modalities reported by the cooperating agents. The concept of uncertainty reasoning framework itself is not a novel idea. The uncertainty frameworks proposed earlier have proven their usefulness of the concept for different applications related to data fusion. However, we think that RSTA missions are unique and the sensors involved in RSTA are different from the ones studied by the data fusion community. We would like to investigate the different uncertainty reasoning models and choose the one closer to a deployable Phase II prototype. The primary military applications for the proposed work include the RSTA program, and other mission planning applications which use reconnaissance data from the Unmanned Air Vehicles. Commercial applications include the medical imaging arena for lesion studies, industrial machine vision, and service robotics.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government