An image deniosing method based on dual-tree complex wavelet transform

作者：Li, Yibing; Zhang, Jing; Ye, Fang^*

来源：Journal of Information and Computational Science, 2014, 11(2): 383-390.

DOI：10.12733/jics20102571

• Bayesian
• Bivariate models
• Directional selectivity
• Dual-tree complex wavelet transform
• REmove noise
• Shift-invariance
• Subbands

The dual-tree complex wavelet transform provides good directional selectivity and shift-invariance. It can describe the details of image better. Since the coefficients of image has strong relativity, better effect can be achieved through adopting statistical method for image processing which sufficiently takes advantage of the relativity between the child and parent coefficients. The image is decomposed using dual-tree complex wavelet transform and combined with bivariate model for the coefficients estimation in high frequency sub-bands. As to the coefficients in the highest frequency sub-band which do not have the parent coefficients, the Bayesian threshold is applied in order to remove noise. The experimental results show the validity of the propose method.

• 出版日期2014
• 单位哈尔滨工程大学