Face recognition using active appearance model

This thesis gives a detailed explanation of the statistical models of shape and texture, the Active Appearance Models (AAM) developed by Cootes et. al, supplies several extensions to basic AAM for better performance and finally uses AAM for face recognition.The primary advantage of AAM is that it uses a priori knowledge of shape and texture and combines this knowledge to form a compact model. This compact model then can be used for various purposes including image segmentation and object detection.The details of face recognition and the results obtained using AAM are given. The recognition accuracy is encouraging in images under various lighting and pose variations.