Institute for Automation and Applied Informatics

Establishment of a new method to interactively segment and visualize complex 3D image data


3D-image stacks with low quality often can not be used to do error-free image processing. A typical example is microscopy of 3D-stacks.

This project supported users in a manual highly accurate segmentation of 3D-stacks. Easy-to-use image processing operators automatically generate possibilities for segment borders in 2D, which can be accepted by a user. These accurate segment borders can be extrapolated in 3D. All extracted segments are then used to generate parametric models of the 3D-segment. Thus, outlier and errors can be identified and suppressed.