抽象的な

Face Detection on Real Life Images and Video Frames Based On Linear Discriminant Analysis

V.Ponmalar, A.Geetha

Face detection system has already become an important issue and several techniques have been developed. It remains as one of the challenging problem in the field of image analysis and computer vision. It also presents a detailed idea about the face detection techniques which improves the rate of detection. Initially the image extracted from video frames, still images and database is to be preprocessed for the normalization of contrast and brightness by using Histogram equalization. Then extract the facial features like eyes, nose and mouth from the facial image using the Linear Discriminant Analysis. Based on the facial feature extracted, the face is detected from the image. Linear Discriminant Analysis mainly performs dimensionality reduction. It also provides us with a small set of features that carry the most relevant information for classification purposes. Linear Discriminant groups the images of the same class and separate images of different classes. LDA performs even well for low resolution images and also some images containing illumination variation and out of plane rotation of faces. Its application is mainly for face recognition and gender classification.