BlueFly Project
3D visualization of SBFSEM image stacks
In order to analyze Serial Block Face Scanning Electron Microscope (SBFSEM) image stacks in all three dimensions and to reconstruct neuronal circuits in the fly optic system, we developed two separate software packages, the 2D and 3D Viewer. Both are implemented mainly in the programming language Java.
2D Viewer
The main task of the software is to extract contour lines from one image and to automatically trace those in the following images (z-direction). A respective algorithm was developed by Jakob Macke and Prof. Bernhard Schölkopf at the MPI for Biological Cybernetics in Tuebingen. The algorithm is implemented in Matlab, the graphical user interface in Java.
The algorithm contains two main steps:
- Preprocessing: Image filtering
- Image segmentation: The computer-aided image segmentation allows to find contour edges and to trace them in the following images. In addition, a manual interaction is possible.
3D Viewer
The main task of the software is to reconstruct the 3D structure. Some parts are implemented in JOGL – Java Bindings for OpenGL.
The picture and movie to the right display 3D surfaces of multiple axonal structures with interlaced SBFSEM images. The area depicted is from the outer chiasm of the blow fly Calliphora vicina.
