Proceedings of the Korean Institute of Information and Commucation Sciences Conference (한국정보통신학회:학술대회논문집)
- 2017.05a
- /
- Pages.793-796
- /
- 2017
A Study on Nonlinear Composit Filter for Mixed Noise Removal
복합 잡음 제거를 위한 비선형 합성 필터에 관한 연구
- Kwon, Se-Ik (Dept. of Control and Instrumentation Eng. Pukyong National University) ;
- Kim, Nam-Ho (Dept. of Control and Instrumentation Eng. Pukyong National University)
- Published : 2017.05.31
Abstract
Image signal can be damaged by a variety of noises during the signal processing, and multiple studies have been conducted to restore these signals. The representative noises to be added in the image are salt and pepper noise, additive white Gaussian noise(AWGN), and the composite noise which two noises are combined. Therefore, the algorithms were proposed to process with quadratic spline interpolation and median filter in case of salt and pepper noise with the central pixel of the local mask, and to process with weight filter by pixel changes in case of AWGN, upon noise determination to restore the damaged image in the composite noise environment, in this article.
영상신호는 신호를 처리하는 과정에서 다양한 잡음에 의해 훼손되어지며, 이러한 신호를 복원하기 위한 많은 연구가 이루어지고 있다. 영상에 첨가되는 잡음에는 salt and pepper 잡음, AWGN(additive white Gaussian noise) 및 두 잡음이 혼합된 복합잡음이 대표적이다. 따라서 본 논문에서는 복합잡음 환경에서 훼손된 영상을 복원하기 위해, 잡음 판단을 통해 국부 마스크의 중심화소가 salt and pepper 잡음인 경우, 2차원 스플라인 보간법 및 메디안 필터로 처리하고, AWGN인 경우, 화소변화에 따른 가중치 필터로 처리하는 알고리즘을 제안하였다.