Steerable Pyramids and Tight Wavelet Frames in L2(Rd)

被引:74
作者
Unser, Michael [1 ]
Chenouard, Nicolas [1 ]
Van De Ville, Dimitri [1 ]
机构
[1] Ecole Polytech Fed Lausanne, Biomed Imaging Grp, CH-1015 Lausanne, Switzerland
基金
瑞士国家科学基金会;
关键词
Directional derivatives; multiresolution decomposition; Riesz transform; steerable filters; steerable pyramid; tight frames; wavelet transform; LINEAR INVERSE PROBLEMS; MULTIRESOLUTION ANALYSIS; STATISTICS; DOMAIN; ENHANCEMENT; RETRIEVAL; ORDER;
D O I
10.1109/TIP.2011.2138147
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
We present a functional framework for the design of tight steerable wavelet frames in any number of dimensions. The 2-D version of the method can be viewed as a generalization of Simoncelli's steerable pyramid that gives access to a larger palette of steerable wavelets via a suitable parametrization. The backbone of our construction is a primal isotropic wavelet frame that provides the multiresolution decomposition of the signal. The steerable wavelets are obtained by applying a one-to-many mapping (N th-order generalized Riesz transform) to the primal ones. The shaping of the steerable wavelets is controlled by an M x M unitary matrix (where M is the number of wavelet channels) that can be selected arbitrarily; this allows for a much wider range of solutions than the traditional equiangular configuration (steerable pyramid). We give a complete functional description of these generalized wavelet transforms and derive their steering equations. We describe some concrete examples of transforms, including some built around a Mallat-type multiresolution analysis of L-2(R-d), and provide a fast Fourier transform-based decomposition algorithm. We also propose a principal-component-based method for signal-adapted wavelet design. Finally, we present some illustrative examples together with a comparison of the denoising performance of various brands of steerable transforms. The results are in favor of an optimized wavelet design (equalized principal component analysis), which consistently performs best.
引用
收藏
页码:2705 / 2721
页数:17
相关论文
共 42 条
[21]   WAVELETS FOR CONTRAST ENHANCEMENT OF DIGITAL MAMMOGRAPHY [J].
LAINE, A ;
FAN, JA ;
YANG, WH .
IEEE ENGINEERING IN MEDICINE AND BIOLOGY MAGAZINE, 1995, 14 (05) :536-550
[22]   MAMMOGRAPHIC FEATURE ENHANCEMENT BY MULTISCALE ANALYSIS [J].
LAINE, AF ;
SCHULER, S ;
FAN, J ;
HUDA, W .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 1994, 13 (04) :725-740
[23]   Image fusion by using steerable pyramid [J].
Liu, Z ;
Tsukada, K ;
Hanasaki, K ;
Ho, YK ;
Dai, YP .
PATTERN RECOGNITION LETTERS, 2001, 22 (09) :929-939
[24]  
Mallat S., 2008, A wavelet Tour of Signal Processing, V3rd
[25]   A THEORY FOR MULTIRESOLUTION SIGNAL DECOMPOSITION - THE WAVELET REPRESENTATION [J].
MALLAT, SG .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1989, 11 (07) :674-693
[26]   Natural image statistics and efficient coding [J].
Olshausen, BA ;
Field, DJ .
NETWORK-COMPUTATION IN NEURAL SYSTEMS, 1996, 7 (02) :333-339
[27]   Nonseparable radial frame multiresolution analysis in multidimensions [J].
Papadakis, M ;
Gogoshin, G ;
Kakadiaris, IA ;
Kouri, DJ ;
Hoffman, DK .
NUMERICAL FUNCTIONAL ANALYSIS AND OPTIMIZATION, 2003, 24 (7-8) :907-928
[28]   Image denoising using scale mixtures of Gaussians in the wavelet domain [J].
Portilla, J ;
Strela, V ;
Wainwright, MJ ;
Simoncelli, EP .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2003, 12 (11) :1338-1351
[29]   A parametric texture model based on joint statistics of complex wavelet coefficients [J].
Portilla, J ;
Simoncelli, EP .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2000, 40 (01) :49-71
[30]   Image denoising in steerable pyramid domain based on a local Laplace prior [J].
Rabbani, Hossein .
PATTERN RECOGNITION, 2009, 42 (09) :2181-2193