In this paper we present an efficient algorithm for the fully automatic creation of tree ring profiles from CT-images that show the cross-section of a piece of wood. This kind of pattern is needed in wood technology for annual ring feature detection in general and in the field of dendrochronology which concerns itself with the dating of past events through the study of tree ring growth.

We propose and compare different image processing approaches for an automatic pith detection using CT-Cross-Section images of logs. Our techniques apply different image preprocessing approaches as well as different actual pith detection methods. Besides acquiring knowledge about the anatomical centre of a log, determination of pith position is often a preliminary step for further automated detection of wood characteristics, like annual ring features or compression wood areas.