3D Hand-Held Surveillance and Real Time Multi-Modal Recognition Device

• NCJ Number: 249535
• Date Published: January 2011
• Author(s): Gregory Steinthal
• Length: 92 pages
• Grant Number(s): 2009-IJ-CX-K001
• Agencies: NIJ-Sponsored
• Publication Type: Test/Measurement; Research (Applied/Empirical); Report (Technical); Report (Grant Sponsored)

Annotation

This project developed a cost-effective, network-centric, mobile, wireless hand-held binocular ID system for law enforcement use in identifying uncooperative persons of interest at ranges of up to 100 meters under uncontrolled lighting and environmental conditions.

Abstract

This system uses 2D interagency facial databases with data-rich 3D facial recognition technology that integrates surveillance operations with biometrics for the identification of persons already in law enforcement facial databases. The project completed its research goals and objectives in developing a fully operational hand-held stereoscopic viewing and biometric face-capturing device. This device was successfully demonstrated and field-tested with a complete end-to-end mobile ID client/server application at ranges up to 100 meters from point of use by the Los Angeles Sheriff's Department. Doubling the 2D facial database size did not significantly impact results. The next step is to further increase performance via improvement to image quality. User guides and manuals are included in this final report. 31 figures and 7 tables