Detection of Human Faces in Video Sequences using NGLBP Feature

Abstract

Machine analysis of face detection is an interesting topic for study in Human-Computer Interaction. The existing studies show that discovering the position and scale of the face region is difficult due to significant illumination variation, noise and appearance variation in unconstrained scenarios. This paper suggests a method to detect the location of face area using recently developed YouTube Video face database. In this work, each frame is formulated by normalization technique and separated into overlapping blocks. The Gabor filter is tuned to extract the Gabor features from the individual blocks. The averaged Gabor features are manipulated and local binary pattern histogram features are extracted. The extracted patterns are passed to the classifier with training images for face region identification. Our experimental results on YouTube video face database exhibits promising results and demonstrate a significant performance improvement when compared to the existing techniques. Furthermore, our proposed work is uncaring to head poses and sturdy to variations in illumination, appearance and noisy images