Abstract
The characterization and estimation of the Hölder regularity of random fields has long been an important topic of Probability theory and Statistics. This notion of regularity has also been widely used in image analysis to measure the roughness of textures. However, such a measure is rarely sufficient to characterize textures as it does not account for their directional properties (e.g., isotropy and anisotropy). In this paper, we present an approach to further characterize directional properties associated with the Hölder regularity of random fields. Using the spectral density, we define a notion of asymptotic topothesy which quantifies directional contributions of field high-frequencies to the Hölder regularity. This notion is related to the topothesy function of the so-called anisotropic fractional Brownian fields, but is defined in a more generic framework of intrinsic random fields. We then propose a method based on multi-oriented quadratic variations to estimate this asymptotic topothesy. Eventually, we evaluate this method on synthetic data and apply it for the characterization of historical photographic papers.
Similar content being viewed by others
References
Abry, P., Clausel, M., Jaffard, S., Roux, S., Vedel, B.: Hyperbolic wavelet transform: an efficient tool for multifractal analysis of anisotropic textures. Rev. Mat. Iberoam 31(1), 313–348 (2015)
Abry, P., Roux, S., Wendt, H., Messier, P., et al.: Multiscale anisotropic texture analysis and classification of photographic prints. IEEE Signal Process. Mag. 32(4), 18–27 (2015)
Adler, R.: The Geometry of Random Fields, vol. 62. SIAM, Philadelphia (2010)
Biermé, H.: Champs aléatoires : autosimilarité, anisotropie et étude directionnelle. Ph.D. thesis, University of Orleans, France (2005)
Biermé, H., Benhamou, C., Richard, F.: Parametric estimation for Gaussian operator scaling random fields and anisotropy analysis of bone radiograph textures. In: Pohl, K. (ed.) Proceedings of MICCAI, pp. 13–24. London (2009)
Biermé, H., Lacaux, C.: Hölder regularity for operator scaling stable random fields. Stoch. Proc. Appl. 119(7), 2222–2248 (2009)
Biermé, H., Meerschaert, M.M., Scheffler, H.P.: Operator scaling stable random fields. Stoch. Proc. Appl. 117(3), 312–332 (2007)
Biermé, H., Moisan, M., Richard, F.: A turning-band method for the simulation of anisotropic fractional Brownian field. J. Comput. Gr. Stat. 24(3), 885–904 (2015)
Biermé, H., Richard, F.: Estimation of anisotropic Gaussian fields through Radon transform. ESAIM Probab. Stat. 12(1), 30–50 (2008)
Biermé, H., Richard, F.: Analysis of texture anisotropy based on some Gaussian fields with spectral density. In: Bergounioux, M. (ed.) Mathematical Image Processing, pp. 59–73. Springer, Berlin (2011)
Bonami, A., Estrade, A.: Anisotropic analysis of some Gaussian models. J. Fourier Anal. Appl. 9, 215–236 (2003)
Chilès, J., Delfiner, P.: Geostatistics: Modeling Spatial Uncertainty, 2nd edn. Wiley, Hoboken (2012)
Clausel, M., Vedel, B.: Explicit construction of operator scaling Gaussian random fields. Fractals 19(01), 101–111 (2011)
Hochmuth, R.: Wavelet characterizations for anisotropic Besov spaces. Appl. Comput. Harmon. Anal. 12(2), 179–208 (2002)
Istas, J., Lang, G.: Quadratic variations and estimation of the local Holder index of a Gaussian process. Ann. Inst. Henri Poincaré 33(4), 407–436 (1997). Prob.Stat
Johnson, R., Messier, P., Sethares, W., et al.: Pursuing automated classification of historic photographic papers from raking light images. J. Am. Inst. Conserv. 53(3), 159–170 (2014)
Mandelbrot, B.B., Van Ness, J.: Fractional Brownian motion, fractional noises and applications. SIAM Rev. 10, 422–437 (1968)
Messier, P., Johnson, R., Wilhelm, H., Sethares, W., Klein, A., et al.: Automated surface texture classification of inkjet and photographic media. In: NIP & Digital Fabrication Conference, pp. 85–91. Society for Imaging Science and Technology (2013)
Richard, F.: Analysis of anisotropic Brownian textures and application to lesion detection in mammograms. Procedia Environ. Sci. 27, 16–20 (2015)
Richard, F.: Some anisotropy indices for the characterization of Brownian textures and their application to breast images. Spat. Stat. 18, 147–162 (2016)
Richard, F.: Tests of isotropy for rough textures of trended images. Stat. Sinica 26(3), 1279–1304 (2016)
Richard, F., Biermé, H.: Statistical tests of anisotropy for fractional Brownian textures. Application to full-field digital mammography. J. Math. Imaging Vis. 36(3), 227–240 (2010)
Roux, S., Clausel, M., Vedel, B., Jaffard, S., Abry, P.: Self-similar anisotropic texture analysis: the hyperbolic wavelet transform contribution. IEEE Trans. Image Process. 22(11), 4353–4363 (2013)
Roux, S., Tremblay, N., Borgnat, P., Abry, P., Wendt, H., Messier, P.: Multiscale anisotropic texture unsupervised clustering for photographic paper. In: IEEE International Workshop on Information Forensics and Security (WIFS), pp. 1–6 (2015)
Slimane, M.B., Braiek, H.B.: Directional and anisotropic regularity and irregularity criteria in Triebel wavelet bases. J. Fourier Anal. Appl. 18(5), 893–914 (2012)
Author information
Authors and Affiliations
Corresponding author
Appendix: Proofs
Appendix: Proofs
Proof
(Proposition 2) For some positive constant c, we have
Since the field density satisfies Condition (6), we further obtain
as A tends to \(+\infty \).
Now, in directions s of the set \(E_0=\{s,\tau ^*(s)>0\}\), we notice that \(\eta (s)=H\) and \(\tau ^*(s)=\tau (s)\). Hence, \(f^*(w)=g_{\tau ,\eta }(w)\) whenever \(\arg (w)\) is in \(E_0\), \(f^*(w)=g_{\tau ,\eta }(w)\), and 0 otherwise. Consequently, in polar coordinate, we have
Then, let us decompose the integral over \(E_0^c\) into the sum of two integrals, one over a set \(F_\delta =\{s, \eta (s)-H>\delta /2 \}\) defined for \(\delta >0\), and the other over a set \(E_\delta =\{s, 0<\eta (s)-H<\delta /2 \}\). It follows that
where \(\mu (F_\delta )\) is the measure of \(F_\delta \) on the unit sphere of \(\mathbb {R}^d\). But, as shown by Richard (2016b), \(\lim _{\delta \rightarrow 0} \mu (F_\delta ) =0\). Hence, letting \(\delta =\log (A)^{1-\alpha }\) for some \(0<\alpha <1\), we obtain
Consequently, \(\Delta _A(x)= o(A^{-2H})\) as A tends to \(+\infty \). \(\square \)
Proof
(Proposition 3) When the increment field is mono-directional, the expression of \(\varGamma _{H,v}\) of Eq. (16) reduces to
The expression of \(\beta _{H,\tau ^*}\) in Eq. (28) follows from the simple coordinate change \(u=\rho \cos (\theta )\). \(\square \)
Proof
(Theorem 2) We aim at estimating the solution \(\tau ^*\) of a linear system \( L_H \, \, \tau = \beta \) with a random vector
where \(B=L^T_H \Sigma ^{-1} L_H\), R is a diagonal matrix, \(\lambda >0\), and \(\tilde{\beta }\) is an unbiased estimate of \(\beta \) of variance \(V(\beta )\).
Since \(\tilde{\beta }\) is unbiased, the expectation of \(\tilde{\tau }_\lambda ^*\) satisfies
Moreover, \(B \tau = L^T_H \Sigma ^{-1} \beta \). Hence,
Thus,
where \(\vert \cdot \vert _2\) denotes the 2-norm. Hence, using norm properties, we get
Therefore,
where \(\nu _-\) is the lowest eigenvalue of B.
Next, we establish a bound for \( \vert (I+\lambda B^{-1} R)^{-1} \vert _2\). For any vector u, we have
Therefore, the lowest eigenvalue of \(B^{-1} R\) is above \(\nu _+^{-1}\). Thus, the one of \(I+\lambda B^{-1} R\) is above \(1+\lambda \nu _+^{-1}\). Consequently,
Using Eq. (30), we eventually obtain the inequality (24).
Now, we turn to the variance of the estimator. We have
Hence,
Moreover, since \(B^{-1}\) is a covariance matrix, any term \(B_{ij}^{-1}\) of \(B^{-1}\) is bounded by \(\sqrt{B_{ii}^{-1}B_{jj}^{-1}}\). Hence, it follows that
where \(\Delta \) is a vector formed by terms \(\sqrt{(B)_{ii}^{-1}}\). Noticing that \(\vert \Delta \vert _2^2=\mathrm {trace}(B^{-1})\) and using Eq. (31), we get
Besides, since \(B \tau ^*= L^T_H \Sigma ^{-1} \beta \), we have
Hence,
Combined with Eqs. (32), (25) follows. \(\square \)
Proof
(Corollary 1) Using expressions of bias and variance in Theorem 2, we clearly see that the function g bounds the relative mean square error. Then, a simple variation analysis of this function suffices to show that it reaches a global minimum at \(\lambda ^*\). \(\square \)
Rights and permissions
About this article
Cite this article
Richard, F.J.P. Anisotropy of Hölder Gaussian random fields: characterization, estimation, and application to image textures. Stat Comput 28, 1155–1168 (2018). https://doi.org/10.1007/s11222-017-9785-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11222-017-9785-z