Abstract
The authors describe a three-dimensional vision sensor which fuses discrete range data obtained by a FG (fiber grating) vision sensor and a 2-D intensity image obtained by a CCD (charge coupled device) camera of the sensor. The range image obtained by the sensor is coarsely sampled data. Therefore, one cannot recognize a 3-D object accurately. On the other hand, though one can get features (such as outlines, edge lines, and vertices) from an intensity image, one cannot determine their locations on 3-D space without some restrictions. To solve these problems, the authors propose a method that can recognize a 3-D object by fusing both information obtained from a discrete range image and an abstracted 2-D image.
| Original language | English |
|---|---|
| Pages | 2368-2372 |
| Number of pages | 5 |
| Publication status | Published - 1991 Dec 1 |
| Externally published | Yes |
| Event | Proceedings of the 1991 International Conference on Industrial Electronics, Control and Instrumentation - IECON '91 - Kobe, Jpn Duration: 1991 Oct 28 → 1991 Nov 1 |
Other
| Other | Proceedings of the 1991 International Conference on Industrial Electronics, Control and Instrumentation - IECON '91 |
|---|---|
| City | Kobe, Jpn |
| Period | 91/10/28 → 91/11/1 |
ASJC Scopus subject areas
- Control and Systems Engineering
- Electrical and Electronic Engineering