Analysis of small infrared target features and learning-based false detection removal for infrared search and track

Abstract

An infrared search and track system is an important research goal for military applications. Although there has been much research into small infrared target detection methods, we cannot apply them in real field situations due to the high false alarm rate caused by clutter. This paper presents a novel target attribute extraction and machine learning-based target discrimination method. In our study, eight target features were extracted and analyzed statistically. Learning-based classifiers, such as SVM and Adaboost, have been incorporated and then compared to conventional classifiers using real infrared images. In addition, the generalization capability has also been inspected for various types of infrared clutter.