Face recognition from videos

Face recognition is an application in computer science, importance and practical usage of which is rapidly increasing day by day. In simplest terms, face recognition aims to build a trained system which recognizes people, who are first introduced to system for training.In this work, a system working on video records is constructed. First, system is trained with videos consisting only one person per each, however having no restriction in pose, angle and rotation. Then different videos with same situations are presented to system for recognition.The basic aim in this study is to analyze; how -instead of using all video frames- using only a subset of frames, automatically selected and preprocessed by a system, affects success rate and processing time.In thesis text, first algorithms for object and face detection, dimension transform and classification -which are Haar-Like Features, Principal Components Analysis & Linear Discriminant Analysis and Support Vector Machines respectively- are introduced. Following that, a system based on these algorithms is constructed and experimental results are presented.