Anders Kaestner's activities

Beam-line activities

Diffential phase contrast imaging: In this project we build equipment to offer differential phase contrast imaging as a permanent part of our instrument repetoir. The new instrument is planned to be implemented until mid 2009.

Instrument performance To measure the performance of the imaging system. One example is to determine the optical resolution of the images. To take this a step further, also to measure the resolution of the reconstructed images. This resolution is based on the resolution of the incoming projections but does also depend on the filter choise of the used algorithm.

Tomography

Framework for CT reconstruction: A module based framework for free configuration of preprocessing and backprojection algorithms. The aim is to add functionality for local tomography, few angle tomography, etc. The design is object oriented and implemented in C++.

A first out-come of this work is that we now can offer reconstruction of samples almost twice as wide as the detector.

Image processing

Applications of scale-space filters in 3D Image enhancement using scale-space or inverse scale-space filters have resulted in improved segmentation performance in several applications. My focus is on the use of these filters with three-dimensional images.

Segmentation Many of the user applications require segmentation of the acquired images. An example is the segmentation of plant roots in two- and three-dimensional images.

Programming

This is a side effect of the previous topics. I enjoy the challenge of squeezing the most performance out of a modern CPU. For 3D image processing and CT applications this is relevant if you don't want sit and wait for the spider web to grow on you...

Contact information