A robust face recognition algorithm for real-world applications [Elektronische Ressource] / von Hazım Kemal Ekenel

A Robust Face RecognitionAlgorithm forReal-World Applicationszur Erlangung des akademischen Grades einesDoktors der Ingenieurwissenschaftender Fakultat fur Informatikder Universitat Fridericiana zu Karlsruhe (TH)genehmigteDissertationvonHaz m Kemal Ekenelaus Samsun, Turk eiTag der mundlichen Prufung: 02.02.2009Erster Gutachter: Prof. Dr. A. WaibelZweiter Prof. Dr. J. KittlerAbstractFace recognition is one of the most challenging problems of computer vision andpattern recognition. The di culty in face recognition arises mainly from facialappearance variations caused by factors, such as expression, illumination, partialface occlusion, and time gap between training and testing data capture. More-over, the performance of face recognition algorithms heavily depends on priorfacial feature localization step. That is, face images need to be aligned verywell before they are fed into a face recognition algorithm, which requires pre-cise facial feature localization. This thesis addresses on solving these two mainproblems |facial appearance variations due to changes in expression, illumina-tion, occlusion, time gap, and imprecise face alignment due to mislocalized facialfeatures| in order to accomplish its goal of building a generic face recognitionalgorithm that can function reliably under real-world conditions.The proposed face recognition algorithm is based on the representation of localfacial regions using the discrete cosine transform (DCT).