This Article 
 Bibliographic References 
 Add to: 
Automated Smoothing of Image and Other Regularly Spaced Data
May 1994 (vol. 16 no. 5)
pp. 460-468

This paper is primarily motivated by the problem of automatically removing unwanted noise from high-dimensional remote sensing imagery. The initial step involves the transformation of the data to a space of intrinsically lower dimensionality and the smoothing of images in the new space. Different images require different amounts of smoothing. The signal (assumed to be mostly smooth with relatively few discontinuities) is estimated from the data using the method of generalized cross-validation. It is shown how the generalized cross-validated thin-plate smoothing spline with observations on a regular grid (in d-dimensions) is easily approximated and computed in the Fourier domain. Space domain approximations are also investigated. The technique is applied to some remote sensing data.

[1] M. Abramowitz and I. A. Stegun,Handbook of Mathematical Functions. New York: Dover, 1965.
[2] M. Berman, "Multidimensional thin-plate smoothing splines: the Fourier approach," CSIRO Division of Mathematics and Statistics Rep. NSW 90/11, 1990.
[3] M. Berman, "Automated smoothing of image data," inProc. Conf. Digital Image Computing: Techniques and Applications, Melbourne, pp. 297-305, 1991.
[4] R. N. Bracewell,The Fourier Transform and Its Applications, 2nd ed. New York: McGraw-Hill, 1986.
[5] M. Buckley, "Fast computation of a discretized thin-plate smoothing spline for image data,"Biometrika, vol. 81, 1994, to appear.
[6] P. Craven and G. Wahba, "Smoothing noisy data with spline functions,"Numer. Math., vol. 31, pp. 377-403, 1979.
[7] O. Dubrule, "Two methods with different objectives: splines and kriging,"Math. Geol., vol. 15, pp. 245-257, 1983.
[8] A. Erdelyi, W. Magnus, F. Oberhettinger, and F. G. Triconi,Tables of Integral Transforms, vol. 2. New York: McGraw-Hill, 1954.
[9] D. Girard, "A fast Monte-Carlo cross-validation procedure for large least squares problems with noisy data,"Namer. Math., vol. 56, pp. 1-23, 1989.
[10] I. S. Gradshteyn and I. M. Ryzhik,Tables of Integrals, Series and Products. New York: Academic Press, 1965.
[11] A. A. Green, M. Berman, P. Switzer, and M. D. Craig, "A transformation for ordering multispectral data in terms of image quality with implications for noise removal,"IEEE Trans. Geosci. Remote Sensing, vol. 26, no. 1, pp. 65-74, 1988.
[12] F. R. de Hoog and M. F. Hutchinson, "An efficient method for calculating smoothing splines using orthogonal transformations,"Numer. Math., vol. 50, pp. 311-319, 1987.
[13] M. F. Hutchinson, "A stochastic estimator of the trace of the influence matrix for Laplacian smoothing splines,"Commun. Statist.-Simul., vol. 19, pp. 433-450.
[14] J. B. Lee, A. S. Woodyatt, and M. Berman, "Enhancement of high spectral resolution remote-sensing data by a noise-adjusted principal components transform,"IEEE Trans. Geosci. Remote Sensing, vol. 28, no. 3, pp. 295-304, 1990.
[15] G. Matheron, "Splines and kriging: Their formal equivalence," Internal Report, Centre de Geostatistique, Ecole des Mines de Paris, Fountainebleau, 1980.
[16] F. O'Sullivan, comments on "Some aspects of the spline smoothing approach to non-parametric regression curve fitting,"J. Roy. Statist. Soc. B, vol. 47, pp. 39-40, 1985.
[17] F. O'Sullivan, "Discretized Laplacian smoothing by Fourier methods,"J. Amer. Statist. Assoc., vol. 86, pp. 634-642, 1991.
[18] S. J. Reeves and R. M. Mersereau, "Optimal estimation of the regularization parameter and stabilising functional for regularized image restoration,"Opt. Eng., vol. 29, no. 5, pp. 446-454, 1990.
[19] R. Sibson and G. Stone, "Computation of thin-plate splines,"SIAM J. Sci. Statist. Comput., vol. 12, pp. 1304-1313, 1991.
[20] B. W. Silverman, "Spline smoothing: The equivalent variable kernel method,"Ann. Statist., vol. 12, pp. 898-916, 1984.
[21] R. C. Singleton, "An algorithm for computing the mixed radix fast Fourier transform,"IEEE Trans. Audio Electroacoust., vol. AU-17, no. 1, June 1969.
[22] M. L. Stein, "A kernel approximation to the kriging predictor of a spatial process,"Ann. Inst. Statist. Math., vol. 43, pp. 61-75, 1991.
[23] A. M. Thompson, J. C. Brown, J. W. Kay, and D. M. Titterington, "A study of methods of choosing the smoothing parameter in image restoration by regularization,"IEEE Trans. Pattern Anal. Machine Intell., vol. 13, pp. 326-339, 1991.
[24] A. M. Thompson, J. W. Kay, and D. M. Titterington, "A cautionary note about crossvalidatory choice,"J. Statist. Comput. Simul., vol. 33, pp. 199-216, 1989.
[25] G. Wahba, "Bayesian confidence intervals for the cross-validated smoothing spline,"J. Roy. Statist. Soc. B, vol. 45, pp. 133-150, 1983.
[26] G. Wahba,Spline Models for Observational Data. Philadelphia: SIAM, 1990.
[27] J. Wendelberger, "The computation of Laplacian smoothing splines with examples," Univ. of Wisconsin, Madison, Tech. Rep. 648, Department of Statistics, 1981.
[28] B. Zhang and M. Stein, "Kernel approximations for universal kriging predictors," preprint.

Index Terms:
remote sensing; Fourier transforms; filtering and prediction theory; image processing; approximation theory; splines (mathematics); high dimensional remote sensing imagery; noise reduction; image smoothing; generalized cross validation; thin plate smoothing spline; Fourier transform; space domain approximations
M. Berman, "Automated Smoothing of Image and Other Regularly Spaced Data," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 16, no. 5, pp. 460-468, May 1994, doi:10.1109/34.291451
Usage of this product signifies your acceptance of the Terms of Use.