Abstract
We know one characteristic of image is that each pixel is very close to surrounding neighbors. If the impulsive noise appears in the local area, it will be very abruptly. In the proposed scheme, we pick up several neighboring pixels around the center pixel in the scanning window as the statistic factors. The mean value and mean of absolute neighboring difference (MAND) are utilized from those. Further, we use them to acquire a critical value. The value is a representative characteristic of those pixels in the local area and it will be used to identify whether if a pixel is corrupted.
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 24th International Conference on Distributed Computing Systems Workshops |
| Editors | N.-F. Tzeng, M. Takizawa |
| Pages | 190-194 |
| Number of pages | 5 |
| Volume | 24 |
| Publication status | Published - 2004 Jul 12 |
| Event | Proceedings - 24th International Conference on Distributed Computing Systems Workshops - Hachioji, Japan Duration: 2004 Mar 23 → 2004 Mar 24 |
Other
| Other | Proceedings - 24th International Conference on Distributed Computing Systems Workshops |
|---|---|
| Country | Japan |
| City | Hachioji |
| Period | 04-03-23 → 04-03-24 |
All Science Journal Classification (ASJC) codes
- Software
- Hardware and Architecture
- Computer Networks and Communications
Fingerprint Dive into the research topics of 'High probability impulsive noise removal for corrupted images'. Together they form a unique fingerprint.
Cite this
Hsiao, J. Y., & Chen, S. Y. (2004). High probability impulsive noise removal for corrupted images. In N-F. Tzeng, & M. Takizawa (Eds.), Proceedings - 24th International Conference on Distributed Computing Systems Workshops (Vol. 24, pp. 190-194)