Dr. Stefan Eickeler

Agnesstr. 65
53225 Bonn

phone: +49(0)228 5366597

Research Interests

Current Projects

Digitization and Management of the Archive of the Neue Zürcher Zeitung (NZZ)

The Neue Zürcher Zeitung (NZZ) is one of Europe's leading and oldest newspaper publishing houses. Its archive contains all of the articles it has ever published, from the first edition (in 1780) until the present day. The archive extends further back in history than almost any other press archive in the world.

Up until now, however, accessing the archive, has been cumbersome. For the period from 1780 to 1871, there is not even an index to help search for articles on particular topics. The archivists have to leaf through every issue one by one, or manually place the microfilms, which have been made of every volume, into the microfilm reader, and then look through them.

On the occasion of its 225th anniversary, and set against this backdrop, the history-conscious publishing house decided to make available its archive in digital form. The digital newspaper archive will make it possible for multiple user groups to perform rapid and precise online searches for relevant newspaper articles.

The newspaper pages are to be presented on the Internet as individual pages, and preferably as full texts. Users enter keywords via an Internet browser and will be able to jump to the relevant position in the digitized newspaper page within seconds.

Two million pages of newsprint documented on 1,500 microfilms will be digitized, and prepared for automatic access. Microfilmed pages whose quality does not suffice for digitization will be re-photographed. The digitization quality will be high enough for both the screen image of the Internet page and for reproductions of newspaper pages.

This task requires the application of the newest technologies in the fields of digitization, image processing and optical character recognition (OCR).

Stefan Betschon. 225 Jahrgäge der NZZ werden elektronisch erschlossen. Neue Zürcher Zeitung, 226(11):59, 14. January 2005

Klaus Jacob. 70 Terabyte Zeitgeschichte. Fraunhofer Magazin, 2.2005, March 2005

Klaus Jacob. 70 terabytes of living history. Fraunhofer Magazin (english version), 2.2005, June 2005

The project NZZ Archive 1780 is the winner of the 2006 Computerworld Honors Program in the category "Media, Arts and Entertainment".

NZZ Archive 1780 Recipient Case Study. The Computerworld Honors Program, 2006

PhD Thesis

Automatische Bildfolgenanalyse mit statistischen Mustererkennungsverfahren

In this thesis new methods for the automatic recognition of the content of image sequences are presented. Solutions to the following video sequences analysis tasks are developed: temporal decomposition of an image sequence into scenes and classification of the scenes, and the recognition of people and their movements in the image sequence. The temporal segmentation of a image sequence and the classification of the segments can be used for image sequences with a given content structure, like broadcast news. The image sequences have a defined chronology of scenes, which belong to certain content classes. The content classes and their chronology are represented by nested Hidden Markov models during the recognition. Another application of the Hidden Markov Modells are the classification of movements of objects in the image sequence. The recognition of human gestures for the application of human-computer-interaction is investigated. The recognition system is capable of recognizing a set of pre-defined gestures that are performed in the viewing area of a camera. The system is able to identify undefined movements and can distinguish them from the gestures. The final task is recognizing people visible in image sequences. The recognition of the people is done by recognizing their faces. The indexing of the faces is composed of the two sub-tasks: detection of the faces and recognition of the faces. It is shown that the face-based video indexing can be used to find known persons in the image sequence as well as to group the people in the sequence unsupervised.

Supervisor: Prof. Gerhard Rigoll, Gerhard-Mercator-University Duisburg

Referee: Prof. Gerhard Rigoll, Gerhard-Mercator-University Duisburg
Referee: Prof. Martin Reiser, Fraunhofer Institut Medienkommunikation, St. Augustin

Keywords: Video Indexing, Face Recognition, Face Detection, Gesture Recognition, Hidden Markov Models

PDF-Document, 141 pages, 3931036 bytes, in German