OpenCV and TYZX : video surveillance for tracking.
As part of the National Security Engineering Institute (NSEI) project, several sensors were developed in conjunction with an assessment algorithm. A camera system was developed in-house to track the locations of personnel within a secure room. In addition, a commercial, off-the-shelf (COTS) tracking system developed by TYZX was examined. TYZX is a Bay Area start-up that has developed its own tracking hardware and software which we use as COTS support for robust tracking. This report discusses the pros and cons of each camera system, how they work, a proposed data fusion method, and some visual results. Distributed, embedded image processing solutions show the most promise in their ability to track multiple targets in complex environments and in real-time. Future work on the camera system may include three-dimensional volumetric tracking by using multiple simple cameras, Kalman or particle filtering, automated camera calibration and registration, and gesture or path recognition.
|Creator/Author:||He, Jim ; Spencer, Andrew ; Chu, Eric|
|Publication Date:||2008 Aug 01|
|OSTI Identifier:||OSTI ID: 942060|
|DOE Contract Number:||AC04-94AL85000|
|Other Number(s):||TRN: US0900380|
|Resource Type:||Technical Report|
|Research Org:||Sandia National Laboratories|
|Subject:||45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEFENSE; CALIBRATION; CAMERAS; IMAGE PROCESSING; NATIONAL SECURITY; PERSONNEL; START-UP; TARGETS|
|Related Subject:||Nuclear facilities-Surveillance operations.; Electronic security systems.; Video.; Security systems.|
|Country of Publication:||United States|
|Format:||Size: 33 p.|
|Update Date:||2009 Feb 02|