Abstract-Band - Fakultät für Informatik, TU Wien - Technische ...

Institut für Rechnergestützte Automation
Arbeitsbereich Computer Vision
Angelika Garz
Efficient Layout Analysis of Ancient Manuscripts Using Local Features
Studium: Masterstudium Medieninformatik
Betreuer: Ao.Univ.Prof. Dr. Robert Sablatnig
A binarization-free layout analysis method for ancient manuscripts is
proposed, which identifies and localizes layout entities exploiting their
structural similarities on the local level. Thus, the textual entities are
disassembled into segments, and a part-based detection is done which
employs local gradient features known from the field of object recognition,
the Scale Invariant Feature Transform (SIFT), to describe these structures. As
the whole entity cannot directly be inferred from the mere positions of the
interest points, a localization algorithm is needed that expands the interest
points according to their scales and the classification score to regions that
encapsulate the whole entity. Hence, a cascading algorithm is proposed that
successively rejects weak candidates applying voting schemes. Layout
analysis is the first step in the process of document understanding; it
identifies regions of interest and hence, serves as input for other algorithms
such as Optical Character Recognition. Moreover, the document layout
allows scholars to establish the spatio-temporal origin, authenticate, or index
a document. The evaluation shows that the method is able to locate main
body text in ancient manuscripts. The detection rate of decorative entities is
not as high as for main body text but already yields to promising results.
Karl-Heinz Nenning
Mirror Visual Feedback Therapy for Phantom Pain: Changes in Functional
Connectivity Patterns
Studium: Masterstudium Medizinische Informatik
Betreuer: Ao.Univ.Prof. Dr. Robert Sablatnig
68
Mirror Visual Feedback Therapy (MVFT) offers efficient non-invasive treatment
for patients suffering from phantom limb pain. It is hypothesized to cause
functional remodeling of neural networks in the patients brain, what induces
a relief in phantom pain. However, details about the functional remodeling of
the brain are not yet fully understood and are a current topic of research. In
this thesis, subject-specific parcellation of functional Magnetic Resonance
Imaging (fMRI) data is utilized and subsequent model map analysis is
employed to quantify changes of functional connectivity patterns related to
MVFT success. Subject-specific functional parcellation is employed in order to
form functionally homogeneous working regions of interest, which adapt to