Two-Dimensional Hermite Filters Simplify the Description of High-Order Statistics of Natural Images

Natural image statistics play a crucial role in shaping biological visual systems, understanding their function and design principles, and designing effective computer-vision algorithms. High-order statistics are critical for conveying local features, but they are challenging to study – largely because their number and variety is large. Here, via the use of two-dimensional Hermite (TDH) functions, we identify a covert symmetry in high-order statistics of natural images that simplifies this task. This emerges from the structure of TDH functions, which are an orthogonal set of functions that are organized into a hierarchy of ranks. Specifically, we find that the shape (skewness and kurtosis) of the distribution of filter coefficients depends only on the projection of the function onto a 1-dimensional subspace specific to each rank. The characterization of natural image statistics provided by TDH filter coefficients reflects both their phase and amplitude structure, and we suggest an intuitive interpretation for the special subspace within each rank.

[1]  Daniel L. Ruderman,et al.  Origins of scaling in natural images , 1996, Vision Research.

[2]  Jean-Bernard Martens.  The Hermite transform-applications , 1990, IEEE Trans. Acoust. Speech Signal Process..

[3]  Siwei Lyu,et al.  Steganalysis using higher-order image statistics , 2006, IEEE Transactions on Information Forensics and Security.

[4]  J. V. van Hateren,et al.  Independent component filters of natural images compared with simple cells in primary visual cortex , 1998, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[5]  Erik Reinhard,et al.  Image Statistics and their Applications in Computer Graphics , 2010, Eurographics.

[6]  C. Zetzsche,et al.  Nonlinear and higher-order approaches to the encoding of natural scenes , 2005, Network.

[7]  Jean-Bernard Martens,et al.  Local orientation analysis in images by means of the Hermite transform , 1997, IEEE Trans. Image Process..

[8]  Jean-Bernard Martens,et al.  The Hermite transform-theory , 1990, IEEE Trans. Acoust. Speech Signal Process..

[9]  D. Ruderman,et al.  Independent component analysis of natural image sequences yields spatio-temporal filters similar to simple cells in primary visual cortex , 1998, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[10]  D. Ruderman.  The statistics of natural images , 1994 .

[11]  A. Réfrégier.  Shapelets: I. a method for image analysis , 2001, astro-ph/0105178.

[12]  José Luis Silván-Cárdenas,et al.  The multiscale Hermite transform for local orientation analysis , 2006, IEEE Transactions on Image Processing.

[13]  Lawrence Sirovich,et al.  The Wigner Transform and Some Exact Properties of Linear Operators , 1982 .

[14]  Siwei Lyu,et al.  Higher-order Wavelet Statistics and their Application to Digital Forensics , 2003, 2003 Conference on Computer Vision and Pattern Recognition Workshop.

[15]  Jonathan D. Victor,et al.  Simultaneously Band and Space Limited Functions in Two Dimensions, and Receptive Fields of Visual Neurons , 2003 .

[16]  D J Field,et al.  Relations between the statistics of natural images and the response properties of cortical cells. , 1987, Journal of the Optical Society of America. A, Optics and image science.

[17]  D. Slepian,et al.  Prolate spheroidal wave functions, fourier analysis and uncertainty — II , 1961 .

[18]  Eero P. Simoncelli,et al.  Nonlinear Extraction of Independent Components of Natural Images Using Radial Gaussianization , 2009, Neural Computation.

[19]  Siwei Lyu,et al.  Detecting Hidden Messages Using Higher-Order Statistics and Support Vector Machines , 2002, Information Hiding.

[20]  Pierre Chainais,et al.  Infinitely Divisible Cascades to Model the Statistics of Natural Images , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[21]  D. Tolhurst,et al.  Amplitude spectra of natural images. , 1992, Ophthalmic & physiological optics : the journal of the British College of Ophthalmic Opticians.

[22]  A.V. Oppenheim,et al.  The importance of phase in signals , 1980, Proceedings of the IEEE.

[23]  J. H. Hateren,et al.  Independent component filters of natural images compared with simple cells in primary visual cortex , 1998 .

[24]  Jean-Bernard Martens,et al.  Image representation and compression with steered Hermite transforms , 1997, Signal Process..

[25]  Eero P. Simoncelli,et al.  A Parametric Texture Model Based on Joint Statistics of Complex Wavelet Coefficients , 2000, International Journal of Computer Vision.

[26]  D. Tolhurst,et al.  Both the phase and the amplitude spectrum may determine the appearance of natural images , 1993, Vision Research.

[27]  R. Baddeley.  Visual perception. An efficient code in V1? , 1996, Nature.

[28]  Matthias Bethge,et al.  Hierarchical Modeling of Local Image Features through $L_p$-Nested Symmetric Distributions , 2009, NIPS.

[29]  Eero P. Simoncelli.  4.7 – Statistical Modeling of Photographic Images , 2005 .

[30]  E. Adelson,et al.  Image statistics and the perception of surface qualities , 2007, Nature.

[31]  Siwei Lyu,et al.  A digital technique for art authentication , 2004, Proc. Natl. Acad. Sci. USA.

[32]  Anjan Chatterjee,et al.  Preference for luminance histogram regularities in natural scenes , 2016, Vision Research.

[33]  Steven W. Zucker,et al.  Understanding the statistics of the natural environment and their implications for vision , 2016, Vision Research.

[34]  D. Burr,et al.  Feature detection in human vision: a phase-dependent energy model , 1988, Proceedings of the Royal Society of London. Series B. Biological Sciences.

[35]  D. Slepian.  Prolate spheroidal wave functions, Fourier analysis and uncertainty — IV: Extensions to many dimensions; generalized prolate spheroidal functions , 1964 .

[36]  M. A. Repucci,et al.  Responses of V1 neurons to two-dimensional hermite functions. , 2006, Journal of neurophysiology.

[37]  Jonathan D. Victor,et al.  Contextual modulation of V1 receptive fields depends on their spatial symmetry , 2009, Journal of Computational Neuroscience.