Feature-based recognition approaches for infrared anti-ship missile seekers

Abstract

Heat seeking missiles pose a major threat to air-, land- and sea-based military platforms, and ongoing research into developing techniques for countering these threats is vital. In order to counter the threat, one needs to understand its performance, by developing high-fidelity models of infrared missile seekers. As seeker technologies advance, the capability exists to include more sophisticated countermeasure rejection techniques, and even techniques to discriminate between different potential targets. This paper considers the application of feature-based ship classification to the acquisition process of an imaging infrared missile in a naval engagement scenario. Scale invariant interest point detectors are used to extract keypoints and descriptors from simulated infrared imagery, generated by a high-fidelity infrared seeker model. Two methods are then used to classify the descriptors into different ship classes: the Generalised Hough Transform for pose estimation, and Bayes Decision Theory using Gaussian mixtures.

Keywords:

• imaging infrared missile
• ship recognition
• scale invariant feature transform
• speeded up robust features
• Gaussian mixtures
• generalised hough transform

Thanks are extended to Chemring Countermeasures who funded this research.