Human Face Detection in Thermal Images Using an Ensemble of Cascading Classifiers

The paper addresses the subject of thermal imagery in the context of face detection. Its aim is to create and investigate a set of cascading classifiers learned on thermal facial portraits. In order to achieve this, an own database was employed, consisting of images from IR thermal camera. Employed classifiers are based on AdaBoost learning method with three types of low-level descriptors, namely Haar-like features, Histogram of oriented Gradients, and Local Binary Patterns. Several schemes of joining classification results were investigated. Performed experiments, on images taken in controlled and uncontrolled conditions, support the conclusions drawn.