摘要
The mutual excitation among the local stimuli satisfying curve distribution (position and orientation continuity) called self-excitation of curves here is an effective method for the discovery and enhancement of visual curves. This article presents a new method using dynamic time-division curve searches and self-excitation. The searches realized by random walks of active particles are guided by inputs, limited by the rules of curve distribution, performed repetitively, and temporally divided for different curve candidates. The time-division operations play an extremely important role in avoiding both the structure division used previously and the global memory of various search routes.
| 源语言 | 英语 |
|---|---|
| 页 | 3500-3504 |
| 页数 | 5 |
| 出版状态 | 已出版 - 1999 |
| 已对外发布 | 是 |
| 活动 | International Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, USA 期限: 10 7月 1999 → 16 7月 1999 |
会议
| 会议 | International Joint Conference on Neural Networks (IJCNN'99) |
|---|---|
| 市 | Washington, DC, USA |
| 时期 | 10/07/99 → 16/07/99 |
学术指纹
探究 'Random time-division operation for salience of visual contours' 的科研主题。它们共同构成独一无二的学术指纹。引用此
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver