We have introduced a geometry which is invariant to certain forms of burning and dodging photographic prints. We then used this geometry to create invariant measurements which represent information which would not change given different photographic printing processes. / The presented algorithm used properties of best-fit planes to represent a photograph. There are many other possibilities for measurements which would fit this framework. Further, the representation of photographs presented in this thesis could be combined with existing computer vision algorithms for such tasks as object recognition within photographs for which we do not know the development process.
| Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.101163 |
| Date | January 2006 |
| Creators | Phillips, Carlos. |
| 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 | © Carlos Phillips, 2006 |
| Relation | alephsysno: 002599952, proquestno: AAIMR32772, Theses scanned by UMI/ProQuest. |
Page generated in 0.0021 seconds