SmartAvoidT - A Portable, Scalable Object Avoidance Solution for all Day/Night/Weather/Smoke Environments

The primary objective of this SBIR effort is to develop a novel object avoidance algorithm dubbed SmartAvoidT. Unmanned vehicles of all types are faced with the common challenge of avoiding objects in the path of the unmanned vehicle. Until a robust, affordable object avoidance solution is provided, true autonomous operation on unmanned vehicles cannot be performed. Genex Technologies, Inc., a 3D and high-speed image processing company proposes to develop and integrate an object avoidance system for USVs. Our focus is to develop a novel object avoidance algorithm called SmartAvoidT that extracts multiple objects/targets out of video/imagery data, establishes individual tracks for each object and maps a path around each object to avoid collisions. The algorithm will then integrate with navigation systems to change course to avoid each object, continuously updating the "mapped" route. The primary innovation of our proposed SmartAvoidT algorithm is the adaptation of a multi-state-machine framework that uses a variety of object detection and target tracking methods to create robust inter-object `awareness' in high noise and cluttered (sea) environments. Our goal is to develop a reliable object avoidance algorithm that can provide robust location and history of the movement of multiple objects in real time within any imagery data (i.e. visible, near infrared, infrared, image intensified (I2), Ultra-Wideband (UWB), or any imagery yielding multiple targets within a stationary or moving background).