Evaluation of infrared missile warning algorithm suites

Abstract

Operation Desert Storm highlighted the effectiveness of infrared (IR) missile warning systems to detect and track hostile missile threats. This, in turn, has sparked renewed interest in the development of IR missile warning algorithms and architectures. The Avionics Directorate has funded several IR missile warning algorithm studies and the performance of six algorithm suites is presented in this paper. All suites were evaluated against registered data from two bands of mid-wave IR (MWIR) imagery and are evaluated for their target detection performance without the aid of tracking algorithms. The effectiveness of each algorithm suite to detect point targets was measured in terms of missed detections and false alarms. The preliminary results show that for targets with low signal to noise ratios (SNR), missile warning algorithm suites which use simple spatial and spectral filters achieve the same performance as more computationally complex algorithm suites.<>