Ruprecht-Karls-Universität Heidelberg
Digital Image Processing -   
Executive summary

Image processing and analysis constitute complex scientific problems. It is all too easy to overlook this complexity because of the enormous capabilities of human vision. The ease with which we can recognize objects even in complex scenes may lead to the erroneous conclusion that computer vision is an easy task. Therefore, successful scientific applications of digital image processing require an intensive interdisciplinary cooperation between the application research area and basic research in image processing itself. The Interdisciplinary Center for Scientific Computing provides the infrastructure to make such interdisciplinary cooperation successful.

From the multitude of activities of the research group, three highlights are pointed out in this executive summary.

1. DFG Research Unit "Image Sequence Processing to Study Dynamical Processes"

The research unit focuses on key scientific problems concerning growth, exchange, and transport processes in environmental and biosciences that can only be resolved by capturing and analyzing image sequences. Funded in December 1995, the research unit could be extended in its second phase that started in March 1998 by two more application areas. Some of the major research results include
- First growth studies of plant leaves with high spatial and temporal resolution (Schmundt et al., 1998)
- First high-resolution concentration field measurements in the aqueous mass boundary layer at free surfaces (Münsterer and Jähne, 1998)
- Measurement of the air-sea gas transfer rate and its mechanisms by active and passive thermography (Jähne and Haußecker, 1998; Schimpf et al., 1999).
- First global maps of tropospheric NO2 concentrations derived form image sequences of the GOME instrument on the ERS2 satellite (Leue et al., 1998)
- Time-resolved analysis and visualization of dynamic processes in living cells and cell nuclei (Tvaruskó, 1999)
- Quantitative analysis of actin filament motility in in vitro motility assays (Uttenweiler et al., 1999)
- A general total least-squares frame work to directly estimate various parameters of dynamically processes such as diffusion coefficients and decay constants in images sequences (Haußecker et al. 1999).

2. "OpenEye" Project"

This project started in 1998 and is funded by the State of Baden-Württemberg with its "Zukunftsiniative Junge Generation". The Institute for Microelectronics Stuttgart (Prof. Höflinger), the Institute for Computer Science V, University of Mannheim (Prof. Männer), and the Research Group Image Processing at the IWR join its expertise for a synergistic approach to develop imaging processing systems for industrial applications including highly dynamic CMOS sensors, fast hardware based on field programmable gate arrays (FPGA), and fast algorithms. Applications include real-time 3-D imaging with hand-held microscopes, fast tracking systems, and on-line control of laser welding processes in cooperation with the companies ATec, Bosch, Eltec, EM Gerätebau, JEL, and Trumpf.

3. The Heidelberg Image Processing Forum

Interdisciplinary research requires a corresponding education for the students. The research group image processing at the IWR is not only responsible for the basic education of physics students in image processing but is also involved in innovative approaches to education. Triggered by discussions between the research group and leaders of the image processing industry, the Heidelberg Image Processing Forum (Heidelberger Bildverarbeitungsforum) was founded in the fall of 1995. Its goal is to foster the interdisciplinary continuous education in the area of digital image processing, to establish and intensify interdisciplinary contacts between university and research institutes in Southern Germany, and to accelerate the transfer of knowledge from research to industrial application.