当前位置: ag平台网址 > ag平台app下载公告 > 正文


学术报告-Image denoising by tensor product complex tight framelets
ag平台网址:2014-04-21  浏览量:

Image denoising by tensor product complex tight framelets韩斌 教授, (加拿大Alberta大学数学与统计科学系) 邀请人:陈迪荣时 间 :2014年4月21日 下午16:00地 点 :数学与系统科学ag平台app下载 学术交流厅主321 报告摘要:Real-valued tensor product (separable) wavelets and framelets are known to have some shortcomings for high dimensional problems such as image processing. For example, they lack directionality and cannot capture edges very well. In this talk, we introduce directional separable complex tight framelets and show that directionality can be greatly improved by using separable complex tight framelets. While keeping the efficient tensor product structure, our approach has the advantages of much better improved directionality and the use of finitely supported complex tight framelets. For the image denoising problem, we show that tensor product complex tight framelets have significant performance gains compared with several state-of-the-art image denoising methods such as undecimated wavelet transform, dual tree complex wavelet transform, shearlets, and etc. 报告人概况:韩斌教授,小波分析领域国际权威专家。1991年毕业于复旦大学,1994年在中科院数学所获硕士学位(导师为龙瑞麟教授),1998年在加拿大Alberta大学获博士学位(导师为贾荣庆教授),后在Princeton大学做博士后(合作导师为I.Daubechies). 目前担任Appl. Computational Harmonic Anal., J. Approx. Theory等刊物编委。

Copyright ©版权所有:ag平台网址 地址:北京市海淀区ag平台app下载路37号      邮编:100191

XML 地图 | Sitemap 地图