Face Detection in the Near-IR Spectrum

Face detection is an important prerequisite step for successful face recognition. Face detection methods reported in the literature are far from perfect and deteriorate ungracefully where lighting conditions cannot be controlled. We propose a method that could potentially outperform state-of-the-art face detection methods in environments with dynamic lighting conditions. The approach capitalizes upon our near-IR skin and face detection methods reported elsewhere [11][12][13]. It ascertains the existence of a face within a skin region by finding the eyes and eyebrows. The eye-eyebrow pairs are determined by extracting appropriate features from multiple near-IR bands. In this paper we introduce a novel feature extraction method we call dynamic integral projection. The method is relatively simple but highly effective because the processing is constrained within the skin region and aided by the near-IR phenomenology.