Southwest Research Institute® (SwRI®) News

Printer Friendly Version

SwRI automated surveillance systems use machine intelligence to increase security effectiveness, alertness

San Antonio, Texas -- March 14, 2003 -- Engineers at Southwest Research Institute® (SwRI®) are applying unique capabilities that can heighten security and improve surveillance for commercial security, motor safety, and homeland security operations. By combining real-time image processing and machine perception to traditional video surveillance methods, the systems can analyze video feeds from multiple surveillance cameras to automatically detect vehicles, packages, and moving objects such as people and animals.

"Current surveillance systems rely on operators to observe multiple monitors and screens for unusual incidents," says Dr. Brent M. Nowak, manager of automation engineering in the SwRI Automation and Data Systems Division. "In these cases, the amount of video data generated can be so overwhelming that the operator could become lax and fail to notice an incident until some time after it occurs, or fail to notice it altogether. The SwRI automated system significantly reduces the volume of data monitored by operators by sounding an alarm only when it automatically detects an incident."

The system, developed using SwRI internal research funds, also provides capabilities not available from other systems, such as instant replay, enabling security personnel to more effectively monitor large areas and handle complex images. Other video motion detection systems use more simplistic approaches to detect incidents and are prone to false triggers from moving foliage, passing headlights, cloud shadows, and more. This limits their effectiveness primarily to highly constrained environments.

"Our system uses algorithms that incorporate temporal processing and model-based analysis so the system recognizes motions normal to that scenario, such as people and vehicles," Nowak says. "The system also disregards or overcomes false triggers caused by moving foliage and shadows and can effectively detect stationary objects left at the scene, such as packages and suitcases."

The SwRI system can be used for:

Perimeter security. Used at large or small facilities, the system can automatically alert security personnel to the presence of people or vehicles.

Interior security. In hallways, corridors, rooms, courtyards, parking areas, and other closed spaces, surveillance cameras can be multiplexed for display only when monitored zones are occupied.

Under-vehicle surveillance. Images of automobile undercarriages can be analyzed to automatically detect attached packages that could contain explosive devices or other contraband.

Motor and pedestrian safety. The presence of large animals or joggers can be automatically detected and combined with systems that would alert passing motorists.

Data acquisition. The system can monitor, quantify, and log pedestrian or vehicle traffic patterns to help optimize the layout of facilities.

The basic system includes cameras, a computer, and SwRI-designed automated video security software that could be used with some existing surveillance systems. Automated surveillance capabilities can be further enhanced with a number of optional components, such as computer-operated pan/tilt/zoom cameras; thermal infrared cameras for night vision and adverse weather conditions; visible or near-infrared illuminators for night vision with conventional cameras; and image intensifiers for long-range night vision with conventional cameras.

Other optional components and applications include long-range infrared/visible wavelength optics for perimeter surveillance of large facilities; remote monitoring of facilities; and transmissions from remote sites over wireless communications networks. The system can also be configured to work with mobile platforms, such as autonomous land or aerial vehicles carrying surveillance cameras and optional analysis computers.

For more information about SwRI's automated surveillance systems, visit videosurveillance.swri.org.

For more information, contact Maria Stothoff, Communications Department, (210) 522-3305, Fax (210) 522-3547, PO Drawer 28510, San Antonio, Texas 78228-0510.

SwRI News separate.gif (834 bytes) SwRI Home