A Novel Multi-Modal, Multi-Sensor General Purpose Processing Engine for Remote Surveillance

This SBIR Phase I project is to develop a novel Smart Optical SensorT (SOS) technology (hardware platform and embedded software) that can be added to any sensor type to remotely process complex algorithms. Existing surveillance systems typically utilize a variety of sensor modalities ranging from acoustic and seismic to video and infrared sensors. Olfactory based sensors and bio/chemical sensors are also becoming popular sensors for the defense of the United States. The algorithms used to process each of the sensor outputs have traditionally been processed on PC-based processors or expensive, large application specific hardware. Such centralized surveillance architecture fails to meet the ever-increasing demands of the military and homeland security for remote tactical surveillance to cover large geographic areas. The SOS concept we propose represents a new paradigm in intelligent surveillance. By incorporating computationally intensive smart processing algorithms into an embedded, low power hardware platform placed at the sensor location, we can eliminate much of the transmission bandwidth requirements demanded by current systems, providing only the data that is desired by the user. The chief goal of SOS research and development is to bring about a tactical network of sensor post-processors that are small enough to be physically added to the sensors to enable remote sensing literally anywhere the user desires.