A Novel Inter-Scale Correlation Image Denoising Method Based on Dual-Tree M-Band Wavelet

摘要：

A novel inter-scale correlation image denoisingmethod based on Dual-tree M-band wavelet(DTT)isproposed in this paper Dual-tree M-band wavelettransform is a shifi-invariant,multi-scale and multi-direction transform based on a Hilbert pair of waveletsinitially proposed by N.Kingsbury.Improving uponXus denosing algorithm based on wavelet inter-scalecorrelation,a new correlation modeling is providedbetween each high frequency detail subimage andcorresponding M subimages in adjacent lowerfrequency scale.In the new algorithm,signal andnoise are distinguished by the strength of thecorrelation,and combined with threshold functions.The experiment result shows that comparing with theclassical denoising methods,for example,waveletdenoising method,Dual-tree complex waveletdenoising method contourlet denoising method and soon the proposed denoising method achieves anexcellent balance between suppressing noise effectivelyand preserving as many image details and edges aspossible.

作者: Jingwen Yah Guide Yang Anfa Zhang

作者单位: Institute of Technology,Shantou University,Shantou,515063,China

会议类型: 国际会议

会议名称: 2008 3rd International Conference on Intelligent System and Knowledge Engineering(第三届智能系统与知识工程国际会议)(ISKE 2008)

会议地点: 厦门

会议语种:英文

页码: 131-136

在线出版日期: 2008-11-17（万方平台首次上网日期，不代表论文的发表时间）