Share Email Print

Proceedings Paper

An adaptive filter technique of image restoration in wavelet domain to reduce ringing artifacts
Author(s): Cheng Yan; Xiangzhong Fang; Yutao Dong
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The paper presents a simple and effective non-iterative method to reduce ringing artifacts of a restored image, which focuses on the artifacts both near the boundaries and close to the edges of the image. We introduce the power spectrum theory into wavelet domain and suppress the ringing artifacts in both the coarse and the fine scales separately. In low frequency subbands, we analyze the power spectrum of low frequency coefficients and calculate the peak regions that cause the ringing artifacts. After determining the peaks, we suppress them by a Center Weighted Median (CWM) filter from outside to inside layer-by-layer. In high frequency subbands, we use an adaptive filter to filter out residual ringing and preserve sharp edges and details. Experimental results demonstrate that our proposed post-filtering algorithm achieves better subjective and objective results compared with other methods. Also, it has very low complexity and high restoration speed.

Paper Details

Date Published: 24 June 2005
PDF: 9 pages
Proc. SPIE 5960, Visual Communications and Image Processing 2005, 59603J (24 June 2005); doi: 10.1117/12.631651
Show Author Affiliations
Cheng Yan, Shanghai Jiaotong Univ. (China)
Xiangzhong Fang, Shanghai Jiaotong Univ. (China)
Yutao Dong, Shanghai Jiaotong Univ. (China)

Published in SPIE Proceedings Vol. 5960:
Visual Communications and Image Processing 2005
Shipeng Li; Fernando Pereira; Heung-Yeung Shum; Andrew G. Tescher, Editor(s)

© SPIE. Terms of Use
Back to Top