Face recognition with artificial neural network

Face recognition is a subject of pattern recognition which has rapidly increasing importance. Face recognition is a task performed by the human brain without much difficulty. The human face has a distinctive features and the human brain uses the visual information of a face as a biometric identifier.The human brain has a complex structure and therefore difficult to analyze. But for the learning, which is the basic function of the brain, simulation can be done. Computer algorithms (Artifcial Neural Networks) have been developed for face recognition that perform the basic functions of the brain by simulation.Face recognition problem can be defined as to identify or verify a person using a database including the collection of known people with a input face pattern. With the advances in technology and increasing need for face recognition applications, researches in this area have increased rapidly in recent years.In this thesis, a face recognition using artificial neural networks has been developed. The gradient pattern of the original image is used in the face recognition process. Some features are derived form the gradient pattern. The face recognition task is performed by using these derived features in the artificial neural network. The tests show that the developed method is succesfull for face recognition.