Tentacle – a multi-camera surveillance and tracking system
Primordial has announced that it has received $848,791 in Air Force Research Laboratory (AFRL) Small Business Innovation Research (SBIR) contracts to develop a multi-camera immersive surveillance system called Tentacle.
Tentacle fuses sensor feeds from fixed cameras, unmanned aerial vehicles (UAVs), unmanned ground vehicles (UGVs), and manned vehicles into a single three-dimensional (3D) display that enables one analyst to effectively monitor dozens of cameras. Tentacle’s 3D display is comprised of imagery, terrain, buildings, and avatars representing entities such as people and vehicles.
Tentacle supports archive queries and user-configured alerts. Tentacle also supports extracting metadata (e.g. person versus vehicle, upper color, and lower color), analyzing behavior (e.g. crossed tripwire or abandoned bag), and performing real-time queries (e.g. highlight people wearing red shirts).
Tentacle works by integrating real-time feature extraction and tagging algorithms for detecting, tracking, and identifying entities with rule-based database query and dynamic alert tools for de-cluttering and illuminating relevant events.
For the effort, Primordial teamed with intuVision, All Hazards Management (AHM), and Carnegie Mellon University (CMU). intuVision is a leader in surveillance camera tracking and alerts. AHM is developing a 3D display for the project. CMU has expertise in tracking vehicles from UAVs such as the Raven.
“This is an exciting and innovative project that enables us to further develop our image and video analysis algorithms,” said Primordial’s president, Randy Milbert . “Analyzing UAV video is a particularly challenging and rewarding. We hope to ultimately provide tools that increase analyst effectiveness and decrease incident response times.”
