Near convex decomposition of polygons for marble thinsection image processing

OData support
Dr. Csorba Kristóf
Department of Automation and Applied Informatics

The thesis was basically made in the project GrainAutLine. The project’s goal is to provide a state-of-the-art thin section image processing graphic interface for geologist students. I implemented one function of this graphic interface, which can decompose a given arbitrary shape to visually convex parts in purpose of the effective processing.

My main task was firstly to search algorithms for convex decomposition and select, and implement the best of them as my opinion. Secondly, I had to integrate the results into the project GrainAutLine. Because of the presence of the big matrices and the enormous number of inequalities, I also had to use a LP (Linear Programming) Solver, that caused me a big challenge.

The program itself was made in the Qt development system on C++ program language. For the image processing, I used the BSD (Berkeley Software Distribution) licensed open-source library the OpenCV. Because of the fact, that more developers were in the team, we also needed to learn the basics of Git handling.


Please sign in to download the files of this thesis.