This thesis focuses on a particular class of representations of visual form, those based on the Blum skeleton [6]. We shall argue that such a representation has a number of desirable properties, which any good shape descriptor should possess. We shall begin this discussion in Section 1.1 but will keep it on a relatively high level leaving the details to Chapter 2. We shall then, in Section 1.2, review some of the previous work on skeletonization and identify the major computational difficulties encountered in the past. Finally, in Section 1.3, we shall briefly discuss how the method presented in this thesis deals with those issues and we shall identify the contributions of this work.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.19679 |
| Date | January 2003 |
| Creators | Dimitrov, Pavel |
| Publisher | McGill University |
| Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
| Language | English |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Format | application/pdf |
| Coverage | Master of Science (School of Computer Science) |
| Rights | All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated. |
| Relation | alephsysno: 002022754, Theses scanned by McGill Library. |
Page generated in 0.0018 seconds