Markerless Motion Analysis in Diffusion Tensor Fields and Its Applications Vom Fachbereich Informatik der Technischen Universität Darmstadt genehmigte DISSERTATION zur Erlangung des akademischen Grades eines Doktor-Ingenieur (Dr.-Ing) von M.E. SANG MIN YOON Geb, in Daejeon, KOREA Referenten der Arbeit: Prof. Dr.-Ing. Dr. h. c. Dr. E. h. José Encarnação . techn. Dieter Fellner Tag der Einreichung: 27. April 2010 Tag der m ündlichen Pr üfung: 28. Juni 2010 Darmst ädter Dissertationen D17 Darmstadt 2010 AbstractThe analysis of deformable objects which have a high-degree of freedom has longbeen encouraged by numerous researchers because it can be applied to such di-verse areas as medical engineering, video surveillance and monitoring, HumanComputer Interaction, browsing of video databases, interactive gaming and othergrowing applications. Within the computerized environments, the systems arelargely separated into marker based motion capture and markerless motion cap-ture. In particular, markerless motion capture and analysis have also been heavilystudied by numerous researchers using local features, color, shape, texture, anddepth map from stereo vision, but it is still a challenging issue in the area of com-puter vision and computer graphics due to partial occlusion, clutter, dependencyof camera viewpoints, high-dimensional state space and pose ambiguity withinthe target object.