Development and Analysis of 2D and 3D Shape Matching and Shape Retrieval Algorithms using Laplace-Beltrami Spectra and Medial Axis Information
Georg Konwisser, Leibniz Universität Hannover,
bachelor thesis
03/2012
Written by Georg Konwisser (gekonwi AT gmail.com):
Abstract
Development of algorithms often focuses on one task: creating an algorithm which is better than another one doing the same task. The developer is thus compelled to answer the question: "Which algorithm is better?" The highly extendable modular framework and application I developed and implemented for this thesis can help developers of any classifying, matching or retrieval algorithms to answer this question with less implementation effort saving development time and cost.
Overview
In this work I will refer to the mentioned framework and application using the current project name AlCom (Algorithm Comparison).
The main outline of this thesis is as follows:
- general introduction to AlCom (chapter 2 and 3)
- important concepts and components used in AlCom which can also be used to reduce the implementation effort of new modules and other AlCom extensions (chapter 2)
- possibilities of customization and extension of AlCom (chapter 2)
- AlCom’s GUI application (chapter 3)
- the implemented extensions in AlCom and the corresponding theoretical basics (chapter 4)
- an analysis of two shape matching algorithms as a demonstration of extension and usage (chapter 5)
Kontakt:
