Acoustic Detection and Verification of Intrusions against Military Facilities

This SBIR project will develop a complete unattended ground sensor (UGS) system for the detection and verification of personnel and vehicle intrusions at military facilities. McQ will build upon its extensive experience with UGS sensors. We will use our OmniSense® UGS system as a platform for the demonstration of this advanced capability. We will build upon the advanced acoustic speech and vehicle detection algorithms that are already a part of this and other McQ UGS systems. We will expand the capabilities of our current algorithms to improve their probability of detection while at the same time improving their ability to reject acoustic signals and noise that are not related to the intended targets. As part of Phase I we will evaluate variants of our speech activity detection schemes that include both energy based and feature based detection methods. We will also evaluate extensions to our fundamental and harmonic vehicle detection methods and also investigate velocity based and spatial based discriminators. Lastly, we will investigate Amry base communications constraints and develop a proposed approach for integrating them into OmniSense®. All of this will result in a special variant of the system that will be production ready by the end of Phase II.