Machine vision enhances modern art

JULY 25, 2007--The documenta (Kassel, Germany) is one of the most important events in modern art. Only handpicked artists exhibit there. One exhibiting artist, Harun Farocki, a Czech artist living in Berlin, is using HALCON machine-vision software from MVTec (Munich, Germany; www.mvtec.com), usually used in industrial environments, to bring his installations to life. "Deep Play" analyzes the 2006 soccer world-championship final (Italy against France) arranged on 12 screens. To implement his artwork, Farocki used automatic image analysis of video sequences of soccer players' trajectories taken from different camera positions.

With HALCON, Farocki's project partner (the chair for image understanding at Technische Universität Munich) visualized the single soccer players' trajectories in a way that never was done before. And so Deep Play shows a new way to bring a soccer game into a viewer's awareness. Many combined angles of views enable new perceptions based on mathematical image analysis.

This project was implemented within the development of automatic methods for game analysis by professors Michael Beetz and Bernd Radig, Technische Universität Munich. Video streams and images of Deep Play can be found on the Internet.