288 IEEE TRANSACTIONS ON IMAGE PROCESSING, VOL. 20, NO. 1, JANUARY 2011 [7] T. T. Nguyen and S. Oraintara, “A class of multiresolution directional filter banks,” IEEE Trans. Signal Process., vol. 55, no. 3, pp. 949–961, Mar. 2007. [8] R. Eslami and H. Radha, “A new family of nonredundant transforms using hybrid wavelets and directional filter banks,” IEEE Trans. Image Process., vol. 16, no. 4, pp. 1152–1167, Apr. 2007. [9] Y. Tanaka, M. Ikehara, and T. Q. Nguyen, “Multiresolution image rep- resentation using combined 2-D and 1-D directional filter banks,” IEEE Trans. Image Process., vol. 18, no. 2, pp. 269–280, Feb. 2009. [10] N. N. Ma, H. K. Xiong, and L. Song, “2-D dual multiresolution decomposition through NUDFB and its application,” in Proc. IEEE 10th Workshop Multimedia Signal Process., Cairns Qld, Oct. 2008, pp. 509–514. [11] Y. P. Lin and P. P. Vaidyanathan, “Theory and design of two-dimen- sional filter banks: A review,” Multidimens. Syst. Signal Process., vol. 7, pp. 263–330, 1996. [12] J. L. Starck, E. J. Candes, and D. L. Donoho, “The curvelet transform for image denoising,” IEEE Trans. Image Process., vol. 11, no. 6, pp. 670–684, Jun. 2002. [13] A. L. Cunha, J. P. Zhou, and M. N. Do, “The nonsubsampled con- tourlet transform: Theory, design, and applications,” IEEE Trans. Image Process., vol. 15, no. 10, pp. 3089–3101, Oct. 2006. [14] K. Guo and D. Labate, “Optimally sparse multidimensional represen- tation using shearlets,” SIAM J. Math. Anal., vol. 9, pp. 298–318, 2007. [15] J. Lim, Two-Dimensional Signal and Image Processing. Upper Saddle River, NJ: Prentice-Hall, 1990. [16] W. S. Lu and R. Mersereau, Two-Dimensional Digital Filters. New York: Marcel Dekker, 1992. [17] P. P. Vaidyanathan, Multirate Systems and Filter Banks. Upper Saddle River: Printice-Hall, 1993. [18] A. Averbuch, R. Coifman, D. Donoho, M. Elad, and M. Israeli, “Fast and accurate polar Fourier transform,” Appl. Comput. Harmon. Anal., vol. 21, pp. 145–167, 2006. [19] A. Averbuch, R. Coifman, D. L. Donoho, M. Israeli, and Y. Shkol- nisky, “A framework for discrete integral transformations I—The pseudopolar Fourier transform,” SIAM J. Sci. Comput., vol. 30, pp. 764–784, 2008. [20] C. Y. F. Ho, B. W. K. Ling, Y. Q. Liu, P. K. S. Tam, and K. L. Teo, “Optimal design of nonuniform FIR transmultiplexer using semi-infi- nite programming,” IEEE Trans. Signal Process., vol. 53, no. 7, pp. 2598–2603, Jul. 2005. [21] C. Y. F. Ho, B. W. K. Ling, Y. Q. Liu, P. K. S. Tam, and K. L. Teo, “Efficient algorithm for solving semi-infinite programming problems and their applications to nonuniform filter bank designs,” IEEE Trans. Signal Process., vol. 54, no. 11, pp. 4223–4232, Nov. 2006. [22] X. M. Xie, S. C. Chan, and T. I. Yuk, “Design of perfect-reconstruction nonuniform recombination filter banks with flexible rational sampling factors,” IEEE Trans. Circuits Syst. I, vol. 52, no. 9, pp. 1965–1981, Sep. 2005. [23] X. M. Xie, S. C. Chan, and T. I. Yuk, “Design of linear-phase recombi- nation nonuniform filter banks,” IEEE Trans. Signal Process., vol. 54, no. 7, pp. 2809–2814, Jul. 2006. [24] T. W. Chen, L. Qiu, and E. W. Bai, “General multirate building struc- tures with application to nonuniform filter banks,” IEEE Trans. Circuits Syst. II, Analog Digit. Signal Process., vol. 45, no. 8, pp. 948–958, Aug. 1998. [25] C. Y. F. Ho, B. W. K. Ling, and P. K. S. Tam, “Representations of linear dual-rate system via single SISO LTI filter, conventional sampler and block sampler,” IEEE Trans. Circuits Syst. II, Exp. Briefs, vol. 55, no. 2, pp. 168–172, Feb. 2008. [26] J. L. Li, T. Q. Nguyen, and S. Tantaratana, “A simple design method for near-perfect-reconstruction nonuniform filter banks,” IEEE Trans. Signal Process., vol. 45, pp. 2105–2109, Aug. 1997. [27] X. Y. Chen, X. M. Xie, and G. M. Shi, “Direct design of near-perfect- reconstruction linear-phase nonuniform filter banks with rational sam- pling factors,” in Proc. IEEE Int. Conf. Acoust., Speech Signal Process., May 2006, pp. 253–256. [28] A. Cohen, I. Daubechies, and J. C. Feauveau, “Biorthogonal bases of compactly supported wavelets,” Commun. Pure Appl. Math., vol. 45, pp. 485–560, 1992. Passive Polarimetric Imagery-Based Material Classification Robust to Illumination Source Position and Viewpoint Thilakam Vimal Thilak Krishna, Charles D. Creusere, Senior Member, IEEE, and David G. Voelz Abstract—Polarization, a property of light that conveys information about the transverse electric field orientation, complements other at- tributes of electromagnetic radiation such as intensity and frequency. Using multiple passive polarimetric images, we develop an iterative, model-based approach to estimate the complex index of refraction and apply it to target classification. Index Terms—Index of refraction estimation, pBRDF modeling, polari- metric imaging. I. INTRODUCTION The polarization signature of light reflected from an object contains information about material composition (dielectric or metal), shape, surface features, and roughness of a surface [1], and it makes sense to exploit such information to enhance the capabilities of optical wave- length remote sensing systems. For this application, a passive system (i.e., the illumination source is the Sun) is often a practical necessity and the challenge is that the polarization signature is a function of both the target material properties and the relative source/target/camera ge- ometry. While there is a fair amount of prior work dealing with mate- rials classification using passively captured polarimetric imagery, most of the approaches use heuristic models and none of them provide clas- sification features that are invariant to the imaging geometry [2]–[8]. Thus, these approaches are not suitable for outdoor remote sensing ap- plications. In contrast to prior work, we propose here to fit a model of the polari- metric generation process that is based upon the underlying reflection physics in order to estimate the intrinsic complex index of refraction of the target material—a quantity that depends solely upon the material and not on the geometry. We use here the polarimetric bidirectional re- flectance distribution function (pBRDF) model developed by Priest and Meier [9], and we apply the well-known Levenberg–Marquardt algo- rithm [10] to solve for the complex index of refraction using the system of nonlinear equations that results from making multiple polarimetric measurements. The problem of estimating the effective index of re- fraction from multiple polarimetric measurements has been previously studied by Jordan, Lewis, and Jakeman [11] with emission data being collected at various angles of incidence to estimate the effective re- fractive index using the principle angle of incidence ellipsometry tech- nique. In contrast, we apply a model-based approach to derive the effec- tive index of refraction and use the result to facilitate material classifica- tion. Jones, Goldstein, and Spaulding have proposed a material classi- Manuscript received August 31, 2009; revised December 15, 2009 and April 07, 2010; accepted May 04, 2010. Date of publication June 10, 2010; date of current version December 17, 2010. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Jenq-Neng Hwang. T. V. Thilak Krishna is with NVIDIA Corporation, Santa Clara, CA 95050 USA. C. D. Creusere and D. G. Voelz are with the Klipsch School of Electrical En- gineering at New Mexico State University, Las Cruces, NM 88003 USA (e-mail: ccreuser@nmsu.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIP.2010.2052274 1057-7149/$26.00 © 2010 IEEE