Abstract
In this paper, a parallel and distributed algorithm on Spiral Architecture for edge- detection is proposed. The proposed algorithm is based on Master-Slave model. The master node uniformly separates an image using a Spiral Multiplication and sends each slave node a sub-image. Each slave node performs edge detection on the sub-image based on the gradient of image brightness function. Image noise is suppressed by a convolution with Gaussian kerne! before edge points are detected. The gradient consists of three components in three diagonal directions. This detection scheme guarantees a well-balancing load among the slave nodes. Its processing speed is greatly reduced through simultaneous and parallel processes on sub-images. Its accuracy is enhanced by a better approximation of the gradient component.
| Original language | English |
|---|---|
| Pages | 186-191 |
| Number of pages | 6 |
| Publication status | Published - 2003 |
| Externally published | Yes |
| Event | Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03 - Las Vegas, NV, United States Duration: 23 Jun 2003 → 26 Jun 2003 |
Conference
| Conference | Proceedings of the International Conference on Imaging Science, Systems and Technology, CISST'03 |
|---|---|
| Country/Territory | United States |
| City | Las Vegas, NV |
| Period | 23/06/03 → 26/06/03 |
Keywords
- Distributed computing
- Distributed processing
- Edge detection
- Parallel processing
- Spiral Architecture
ASJC Scopus subject areas
- General Engineering