Adaptive threat detection and classification from dual band IR images

Ground combat vehicles are threatened by a variety of guided and smart munitions. Many of these munitions overmatch practical armor. Because of this, modern CONOPS consists of a layered approach to vehicle protection. One of these layers may be stated as: “don´t be hit.” Since, vehicle maneuvering is usually insufficient to meet this requirement, hit-avoidance suites are now under development. An essential component of a hit-avoidance suite is threat warning. The threat warning component must detect, locate, and identify threats in time to effectively apply countermeasures. We present a practical, low cost, dual band IR threat warning system. The system consists of frame registration and threat detection, followed by feature extraction and classification. The system is designed for real time implementation. System performance is shown for real threats measured in real environments