Historical Document Analysis

Scanned documents are a rich source of various information that can be processed utilizing a document analysis system. Such a system covers the areas of Machine Learning, Computer Vision and Natural Language Processing. In the thesis, these areas are covered with a focus on common and state-of-the-art approaches applicable to historical document analysis, which is still challenging due to several difficulties such as handwritten text. Finally, the current research results and aims of the future doctoral thesis are presented