한국컴퓨터산업교육학회:학술대회논문집
- 한국컴퓨터산업교육학회 2003년도 제4회 종합학술대회 논문집
- /
- Pages.87-94
- /
- 2003
평균곡률 확산을 이용한 잡음감소 기법
Noise reduction method using mean curvature diffusion
- Ye Chul-Soo (Dept. of information & Telecommunication, Far East University) ;
-
Chung Hun-Suk
(Dept. of information & Telecommunication, Far East University) ;
- Kim Seong-Jong (Dept. of information & Telecommunication, Far East University) ;
- Hyun Deuk-Chang (Dept. of information & Telecommunication, Far East University)
- 발행 : 2003.11.01
초록
Anisotropic diffusion is a selective smoothing technique that promotes smoothing within a region instead of smoothing across boundaries. In anisotropic diffusion, the rate of smoothing is controlled by the local value of the diffusion coefficient chosen to be a function of the local image gradient magnitude. El-Fallah and Gary E. Ford represented the image as a surface and proved that setting the inhomogeneous diffusion coefficient equal to the inverse of the magnitude of the surface normal results in surface evolving speed that is proportional to the mean curvature of the image surface. This model has the advantage of having the mean curvature diffusion (MCD) render invariant magnitude, thereby preserving structure and locality. In this paper, the proposed MCD model efficiently reduces diffusion coefficient at the thin edges using the smoothness of the surface.
키워드