The analysis of cells and their development and behavior under different conditions is of major importance in biotechnology. However, the necessity for human observation is a major drawback. Cell cultures are negatively influenced when they are moved around, and manual counting of individual cells is tedious and error-prone.

Automatic methods to extract the necessary information could be of immense benefit. The low contrast and high number of shape types that occur in cell imaging, plus the artifacts due to microscopic visualization is a major challenge though.

The seminar will cover different kinds of image processing and pattern recognition methods that are used for automatic cell image analysis.

Content

Topics might include, but are not limited to (students are welcome to suggest own topics):

• Different microscopes and images generated with them
• Basic image processing techniques for cell image enhancement
• Autofocus
• Pixel-/Region-/Model-based segmentation techniques for cell segmentation
• Clustering and classification methods
• ...   

The individual projects mostly comprise a presentation of the theoretical background, plus a programming assignment.