基于脊波变换的体数据压缩编码方法
|
摘要
体数据的数据量大、数据间的相关性强、拥有大量的线或面结构,因此需要研究有效的压缩编码方法。脊波变换作为一种新的时频分析工具,在处理线或面的奇异性时有它适用的一面。在介绍脊波变换理论的基础上,将脊波变换的思想应用到体数据的压缩编码中。文中两种压缩策略的主要思想分别为:策略1先将体数据划分成切片组,再对每一张切片做二维脊波变换,然后进行量化和熵编码;策略2直接对体数据做类似于三维脊波变换的变换,然后进行量化和熵编码。比较而言,策略1实现简单,策略2能获得更高的压缩比。两种策略都具有较强的鲁棒性,且能实现嵌入式编码。该方法已应用到实际工业CT体数据的压缩编码中,还可用于其它类型体数据的压缩编码中。
The cubic data are enormous,the relationship among them are strong,and there are lots of singularities as line or plane,so we need an effective compression coding method.The ridgelet,as a new analytic tool,has the good performance for describing the signals which have super-plane singularities in high dimensions.In this paper,on the basis of the theory of the ridgelet transform,two kind of compression strategies are adopted to the cubic data compression coding.The first one is the idea of the 2D ridgelet transform applied on each slice of the cubic data,the second one is the idea of the similar 3D ridgelet transform applied on the whole cubic data directly.Comparatively speaking,the first one is simpler and the second one has more compression ability.Though each of the two strategies have its own strong point,they have the same characteristics:strong robustness and embedded coding.Our methods have been applied to the cubic data in ICT,they can be applied to other cubic data too.
The cubic data are enormous,the relationship among them are strong,and there are lots of singularities as line or plane,so we need an effective compression coding method.The ridgelet,as a new analytic tool,has the good performance for describing the signals which have super-plane singularities in high dimensions.In this paper,on the basis of the theory of the ridgelet transform,two kind of compression strategies are adopted to the cubic data compression coding.The first one is the idea of the 2D ridgelet transform applied on each slice of the cubic data,the second one is the idea of the similar 3D ridgelet transform applied on the whole cubic data directly.Comparatively speaking,the first one is simpler and the second one has more compression ability.Though each of the two strategies have its own strong point,they have the same characteristics:strong robustness and embedded coding.Our methods have been applied to the cubic data in ICT,they can be applied to other cubic data too.
引文
[1]Mallat S.A wavelet tour of signal processing[M].北京:机械工业出版社,1998.
[2]ZENG Li,Jansen C P,Marsch S,et al.Wavelet compression of four-dimensional arbitrarily size echocardiographic data[J].IEEE Trans on Medical Imaging,2002,21(9):1179-1187.
[3]Candes E J.Ridgelet:theory and application[D].Department of Statis-tics,Stanford University,1998.
[4]Candes E J,Donoho D L.Ridgelet:a key to higher-dimensional in-termittency?[J].Philosophical Transactions of the Royal Society of London.Ser.A,1999,357:2495-2507.
[5]Do M N,Vetterli M.Orthornormal finite ridgelet transform for im-age compression[C]//ICIP.New York:IEEE,2000:367-370.
[6]Granai L,Moschetti F,Vanderghevnst P.Ridgelet transform applied to motion compensated images[C]//ICASSP.New York:IEEE,2003,3:381-384.
[7]袁修贵,宋守根.Ridgelet变换在地震数据压缩中的应用[J].中南大学学报:自然科学版,2004,35(2):300-303.
[8]袁修贵,张安.Ridgelet变换在图像压缩中的应用[J].计算机工程与科学,2005,27(11):34-36.
[9]张选德,宋国乡.基于脊波变换的图像压缩算法[J].现代电子技术,2006(6):62-64.
[10]Beylkin G.Discrete Radon transform[J].IEEE Trans ASSP,1987,35(1):162-172.
[11]HU JICUN.Cone beam tomographic image reconstruction:algorithm development and application in small animal imaging[D];Depart-ment of Biomedical Engineering,Maruette University;Department of Biophysics,Medical College of Wisconsin,2004.
[12]张旭东,卢国栋,冯健.图像编码基础和小波压缩技术——原理、算法和标准[M].北京:清华大学出版社,2004.
[13]http://www.gris.uni-tuebingen.de/areas/scivis/volren/datasets/data/en-gine.raw.gz.
[2]ZENG Li,Jansen C P,Marsch S,et al.Wavelet compression of four-dimensional arbitrarily size echocardiographic data[J].IEEE Trans on Medical Imaging,2002,21(9):1179-1187.
[3]Candes E J.Ridgelet:theory and application[D].Department of Statis-tics,Stanford University,1998.
[4]Candes E J,Donoho D L.Ridgelet:a key to higher-dimensional in-termittency?[J].Philosophical Transactions of the Royal Society of London.Ser.A,1999,357:2495-2507.
[5]Do M N,Vetterli M.Orthornormal finite ridgelet transform for im-age compression[C]//ICIP.New York:IEEE,2000:367-370.
[6]Granai L,Moschetti F,Vanderghevnst P.Ridgelet transform applied to motion compensated images[C]//ICASSP.New York:IEEE,2003,3:381-384.
[7]袁修贵,宋守根.Ridgelet变换在地震数据压缩中的应用[J].中南大学学报:自然科学版,2004,35(2):300-303.
[8]袁修贵,张安.Ridgelet变换在图像压缩中的应用[J].计算机工程与科学,2005,27(11):34-36.
[9]张选德,宋国乡.基于脊波变换的图像压缩算法[J].现代电子技术,2006(6):62-64.
[10]Beylkin G.Discrete Radon transform[J].IEEE Trans ASSP,1987,35(1):162-172.
[11]HU JICUN.Cone beam tomographic image reconstruction:algorithm development and application in small animal imaging[D];Depart-ment of Biomedical Engineering,Maruette University;Department of Biophysics,Medical College of Wisconsin,2004.
[12]张旭东,卢国栋,冯健.图像编码基础和小波压缩技术——原理、算法和标准[M].北京:清华大学出版社,2004.
[13]http://www.gris.uni-tuebingen.de/areas/scivis/volren/datasets/data/en-gine.raw.gz.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心