Name
Abstract | ||
|---|---|---|
Petroglyphs can be found on rock panels all over the world. The possibilities of digital photography and more recently various 3D scanning methods opened a new stage for the documentation and analysis of petroglyphs. The existing work on petroglyph shape similarity has largely avoided the questions of articulation, merged petroglyphs and potentially missing parts of petroglyphs. We aim at contributing to close this gap by applying a novel petroglyph shape descriptor based on the skeletal graph. Our contribution is twofold: First, we provide a real-world dataset of petroglyph shapes. Second, we propose a graph-based shape descriptor for petroglyphs. Comprehensive evaluations show, that the combination of the proposed descriptor with existing ones improves the performance in petroglyph shape similarity modeling. |
Year | DOI | Venue |
|---|---|---|
2014 | 10.1007/978-3-319-16178-5_9 | COMPUTER VISION - ECCV 2014 WORKSHOPS, PT I |
Keywords | Field | DocType |
Petroglyph similarity, Shape similarity, Graph matching, Graph edit distance, Graph embedding | Computer vision,Graph,Digital photography,Computer graphics (images),Computer science,Graph embedding,Matching (graph theory),Artificial intelligence,Documentation,Petroglyph,Graph edit distance | Conference |
Volume | ISSN | Citations |
8925 | 0302-9743 | 2 |
PageRank | References | Authors |
0.37 | 33 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Markus Seidl | 1 | 30 | 4.36 |
| Ewald Wieser | 2 | 8 | 1.17 |
| Matthias Zeppelzauer | 3 | 186 | 21.35 |
| Axel Pinz | 4 | 1632 | 112.96 |
| Christian Breiteneder | 5 | 410 | 288.17 |