A face detection and recognition system based on rectangular feature orientation

We propose a face detection and recognition system based on the Adaboost fast face detection technique. This system incorporates classifiers of eyes and lips to detect the eyes and lips regions. These regions (eyes and lips) can be used to locate a proper region for nose. With the aid of these facial features and the relative distance measures of these individual regions, an artificial neural network can be trained and then be used as a face recognizer. The proposed method has been implemented successfully by using the OpenCV. Experimental results show that our method is valid and efficient.