Pose estimation from a two-dimensional view by use of composite correlation filters and neural networks.

We present a technique to estimate the pose of a three-dimensional object from a two-dimensional view. We first compute the correlation between the unknown image and several synthetic-discriminant-function filters constructed with known views of the object. We consider both linear and nonlinear correlations. The filters are constructed in such a way that the obtained correlation values depend on the pose parameters. We show that this dependence is not perfectly linear, in particular for nonlinear correlation. Therefore we use a two-layer neural network to retrieve the pose parameters from the correlation values. We demonstrate the technique by simultaneously estimating the in-plane and out-of-plane orientations of an airplane within an 8-deg portion. We show that a nonlinear correlation is necessary to identify the object and also to estimate its pose. On the other hand, linear correlation is more accurate and more robust. A combination of linear and nonlinear correlations gives the best results.

[1]  Gregory D. Hager,et al.  Fast and Globally Convergent Pose Estimation from Video Images , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  B Javidi,et al.  Nonlinear joint-transform correlation: an optimal solution for adaptive image discrimination and input noise robustness. , 1994, Optics letters.

[3]  T.,et al.  Training Feedforward Networks with the Marquardt Algorithm , 2004 .

[4]  Abhijit Mahalanobis.  Correlation filters for object tracking, target reacquisition, and smart aim-point selection , 1997, Defense, Security, and Sensing.

[5]  Norberto F. Ezquerra,et al.  An approach to 3D pose determination , 1996, TOGS.

[6]  H J Caulfield,et al.  Improved discrimination in optical character recognition. , 1969, Applied optics.

[7]  F Dubois.  Nonredundant filters for pattern recognition and orientation estimation of rotated images. , 1991, Applied optics.

[8]  Laurence G. Hassebrook,et al.  Postprocessing of correlation for orientation estimation , 1997 .

[9]  Koichiro Deguchi,et al.  Head pose determination from one image using a generic model , 1998, Proceedings Third IEEE International Conference on Automatic Face and Gesture Recognition.

[10]  Ryutarou Ohbuchi,et al.  Merging virtual objects with the real world: seeing ultrasound imagery within the patient , 1992, SIGGRAPH.

[11]  Bahram Javidi,et al.  Distortion-invariant pattern recognition with Fourier-plane nonlinear filters. , 1996, Applied optics.

[12]  D Casasent,et al.  Multivariant technique for multiclass pattern recognition. , 1980, Applied optics.

[13]  William L. Wilson,et al.  VISUAL SERVO CONTROL OF ROBOTS USING KALMAN FILTER ESTIMATES OF ROBOT POSE RELATIVE TO WORK-PIECES , 1993 .

[14]  Christophe Chesnaud,et al.  Statistical Region Snake-Based Segmentation Adapted to Different Physical Noise Models , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Shimon Ullman,et al.  Recognizing solid objects by alignment with an image , 1990, International Journal of Computer Vision.

[16]  Linda G. Shapiro,et al.  Image Segmentation Techniques , 1984, Other Conferences.

[17]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[18]  Xie Yuan-dan,et al.  Survey on Image Segmentation , 2002 .

[19]  Larry S. Davis,et al.  Computing 3-D head orientation from a monocular image sequence , 1996, Proceedings of the Second International Conference on Automatic Face and Gesture Recognition.

[20]  Larry S. Davis,et al.  Computing 3-D head orientation from a monocular image sequence , 1996, Proceedings of the Second International Conference on Automatic Face and Gesture Recognition.

[21]  Bahram Javidi,et al.  Optimum distortion-invariant filter for detecting a noisy distorted target in nonoverlapping background noise , 1995 .

[22]  B. Javidi.  Nonlinear joint power spectrum based optical correlation. , 1989, Applied optics.

[23]  D Casasent,et al.  Unified synthetic discriminant function computational formulation. , 1984, Applied optics.

[24]  Robert M. Haralick,et al.  Review and analysis of solutions of the three point perspective pose estimation problem , 1994, International Journal of Computer Vision.

[25]  Richard D. Juday,et al.  Multidimensional synthetic estimation filter , 1990, Optics & Photonics.

[26]  B. V. K. Vijaya Kumar,et al.  Correlation Filters For Orientation Estimation , 1988, Defense, Security, and Sensing.

[27]  FuchsHenry,et al.  Merging virtual objects with the real world , 1992 .

[28]  H J Caulfield.  Linear combination of filters for character recognition: a unified treatment. , 1980, Applied optics.

[29]  Abhijit Mahalanobis,et al.  Review of correlation filters and their application for scene matching , 1996, Optics + Photonics.

[30]  Bahram Javidi,et al.  Optimum nonlinear composite filter for distortion-tolerant pattern recognition. , 2002, Applied optics.

[31]  Haiyuan Wu,et al.  Face detection and rotations estimation using color information , 1996, Proceedings 5th IEEE International Workshop on Robot and Human Communication. RO-MAN'96 TSUKUBA.