Abstract
Binary direction relations between points in two-dimensional space are the basis to any qualitative direction calculus. Previous calculi are only on a very low level of granularity. In this paper we propose a generalization of previous approaches which enables qualitative calculi with an arbitrary level of granularity. The resulting calculi are so powerful that they can even emulate a quantitative representation based on a coordinate system. We also propose a less powerful, purely qualitative version of the generalized calculus. We identify tractable subsets of the generalized calculus and describe some applications for which these calculi are useful.
| Original language | English |
|---|---|
| Pages (from-to) | 65-74 |
| Number of pages | 10 |
| Journal | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
| Volume | 3157 |
| DOIs | |
| Publication status | Published - 2004 |
| Externally published | Yes |
| Event | 8th Pacific Rim International Conference on Artificial Intelligence, PRICAI 2004: Trends in Artificial Intelligence - Auckland, New Zealand Duration: 9 Aug 2004 → 13 Aug 2004 |
Fingerprint
Dive into the research topics of 'Qualitative direction calculi with arbitrary granularity'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver