KBR Wyle Services, LLC — Department of Defense SBIR Phase I: AF151-017

Description

ABSTRACT:The Design Knowledge Companys (TDKCs) proposed Technique for a Robust, Autonomous CPOHT System (TRACS) product combines innovative image feature extraction, registration, tracking, selection, integration, and helmet state calculation processes to achieve this research objective. TRACS includes four principal components: (1) passive high-resolution cameras, placement and number of cameras to be specified in Phase I; (2) a feature extractor which uses fast-executing image registration and object location/tracking image processing algorithm techniques; (3) a feature selector and integrator whose image processing algorithms pre-process the raw position data to reduce, integrate, and select features from the multiple cameras for optimal accuracy and speed; and (4) a helmet state calculator which produces and tracks the full, six-degrees-of-freedom (6-DOF) helmet position state in three-dimensional (3-D) space coordinates. The TDKC team is well-qualified to perform this research due to our related experience, success with similar developments, and proven expertise with sensing, video/image processing, and real-time computing technologies. During Phase I, we will build, test, and demonstrate a TRACS Feasibility Testbedan early functional prototype to experimentally and analytically assess feasibility and to verify and demonstrate our feasibility conclusions to the Air Force.BENEFIT:TRACS integrates these innovative image processing technologies to provide accurate 6-DOF helmet tracking at real-time speed, using TDKCs existing and proven video-speed image registration technique, the Simple Topographical Area Registration Engine or STARE. TRACS provides this helmet tracking in a small-volume, transportable package with minimal setup or calibration requirements, and autonomous operation. TRACS offers cost-effective and accurate passive helmet tracking advantages to numerous government or military organizations that need to track helmet (or other object) dynamic 6-DOF state in real-time with high accuracy. Commercial applications are significant, predominantly in the gaming and security domains, as well as a lower-cost, higher-speed alternative to tracking or characterizing object motion in general, which may occur in reconnaissance, surveillance, or other missions.