On Kernel Selection of Multivariate Local Polynomial Modelling and its Application to Image Smoothing and Reconstruction

This paper studies the problem of adaptive kernel selection for multivariate local polynomial regression (LPR) and its application to smoothing and reconstruction of noisy images. In multivariate LPR, the multidimensional signals are modeled locally by a polynomial using least-squares (LS) criterion with a kernel controlled by a certain bandwidth matrix. Based on the traditional intersection confidence intervals (ICI) method, a new refined ICI (RICI) adaptive scale selector for symmetric kernel is developed to achieve a better bias-variance tradeoff. The method is further extended to steering kernel with local orientation to adapt better to local characteristics of multidimensional signals. The resulting multivariate LPR method called the steering-kernel-based LPR with refined ICI method (SK-LPR-RICI) is applied to the smoothing and reconstruction problems in noisy images. Simulation results show that the proposed SK-LPR-RICI method has a better PSNR and visual performance than conventional LPR-based methods in image processing.

[1]  Shmuel Peleg,et al.  Improving image resolution using subpixel motion , 1987, Pattern Recognit. Lett..

[2]  P. Vieu.  Nonparametric Regression: Optimal Local Bandwidth Choice , 1991 .

[3]  Peyman Milanfar,et al.  Kernel Regression for Image Processing and Reconstruction , 2007, IEEE Transactions on Image Processing.

[4]  Bruce H. Pillman.  Super-Resolution Imaging , 2013, J. Electronic Imaging.

[5]  Elias Masry,et al.  MULTIVARIATE LOCAL POLYNOMIAL REGRESSION FOR TIME SERIES:UNIFORM STRONG CONSISTENCY AND RATES , 1996 .

[6]  M. Wand,et al.  An Effective Bandwidth Selector for Local Least Squares Regression , 1995 .

[7]  Jaakko Astola,et al.  Local Approximation Techniques in Signal and Image Processing (SPIE Press Monograph Vol. PM157) , 2006 .

[8]  Michal Irani,et al.  Improving resolution by image registration , 1991, CVGIP Graph. Model. Image Process..

[9]  Lijian Yang,et al.  Multivariate bandwidth selection for local linear regression , 1999 .

[10]  Michael Elad,et al.  Fast and robust multiframe super resolution , 2004, IEEE Transactions on Image Processing.

[11]  Jianqing Fan,et al.  Data‐Driven Bandwidth Selection in Local Polynomial Fitting: Variable Bandwidth and Spatial Adaptation , 1995 .

[12]  Jaakko Astola,et al.  A spatially adaptive nonparametric regression image deblurring , 2005, IEEE Transactions on Image Processing.

[13]  Michael Elad,et al.  Multiframe demosaicing and super-resolution of color images , 2006, IEEE Transactions on Image Processing.

[14]  M. Wand,et al.  Multivariate Locally Weighted Least Squares Regression , 1994 .

[15]  Ljubisa Stankovic,et al.  Performance analysis of the adaptive algorithm for bias-to-variance tradeoff , 2004, IEEE Transactions on Signal Processing.

[16]  Shing-Chow Chan,et al.  On Bandwidth Selection in Local Polynomial Regression Analysis and Its Application to Multi-resolution Analysis of Non-uniform Data , 2008, J. Signal Process. Syst..

[17]  Elias Masry,et al.  Multivariate regression estimation: local polynomial fitting for time series , 1997 .

[18]  M. C. Jones,et al.  On optimal data-based bandwidth selection in kernel density estimation , 1991 .

[19]  Jianqing Fan,et al.  Local polynomial modelling and its applications , 1994 .

[20]  O. Lepskii.  On a Problem of Adaptive Estimation in Gaussian White Noise , 1991 .

[21]  A. Goldenshluger.  On Spatial Adaptive Estimation of Nonparametric Regression , 2004 .

[22]  Shing-Chow Chan,et al.  Multi-resolution analysis of non-uniform data with jump discontinuities and impulsive noise using robust local polynomial regression , 2004, 2004 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[23]  Irène Gijbels,et al.  A study of variable bandwidth selection for local polynomial regression , 1996 .

[24]  Shing-Chow Chan,et al.  Robust local polynomial regression using M-estimator with adaptive bandwidth , 2004, 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512).

[25]  D. Ruppert.  Empirical-Bias Bandwidths for Local Polynomial Nonparametric Regression and Density Estimation , 1997 .

[26]  Vladimir Katkovnik,et al.  A new method for varying adaptive bandwidth selection , 1999, IEEE Trans. Signal Process..

[27]  Elias Masry,et al.  Multivariate regression estimation of continuous-time processes from sampled data: Local polynomial fitting approach , 1999, IEEE Trans. Inf. Theory.

[28]  Shing-Chow Chan,et al.  A new two-stage method for restoration of images corrupted by Gaussian and impulse noises using local polynomial regression and edge preserving regularization , 2009, 2009 IEEE International Symposium on Circuits and Systems.

[29]  Matthew P. Wand,et al.  Kernel Smoothing , 1995 .

[30]  M. C. Jones,et al.  A reliable data-based bandwidth selection method for kernel density estimation , 1991 .

[31]  Sabine Süsstrunk,et al.  A Frequency Domain Approach to Registration of Aliased Images with Application to Super-resolution , 2006, EURASIP J. Adv. Signal Process..

[32]  M. Loew,et al.  Closed-form quality measures for compressed medical images: statistical preliminaries for transform coding , 2003, Proceedings of the 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (IEEE Cat. No.03CH37439).