In the case of serious bone injuries, implantation of different so-called grafts into the damaged area is an often employed method in order to advance the bone regeneration.
There are special experiments at the Semmelweis University to examine the usability of different implants. At the Institute of Physiology a research group has been engaged with the topic. The aim of this investigation is to examine the growing difference of different bones graft.
The separation of the different bone regions is a difficult task of this examination, it is based on the investigation of the bone’s structure. Developing an automated method, that is able to separate automatically the regions with different bone structure is necessary for executing the large number of experiments.
The main goal of my work was the development and validation of an efficient method for the separation of the animal and human bone region.
This diploma thesis gives brief overview of the different image processing methods. After the introduction of the various techniques, I present the selection of the most efficient one to solve the given problem. Finally I implemented the developed algorithm, and tested it on real bone images. The development work and evaluation of the results is presented in the second part of this thesis.