
This Article  
 
Share  
Bibliographic References  
Add to:  
Digg Furl Spurl Blink Simpy Del.icio.us Y!MyWeb  
Search  
 
ASCII Text  x  
J. Konrad, E. Dubois, "Bayesian Estimation of Motion Vector Fields," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 14, no. 9, pp. 910927, September, 1992.  
BibTex  x  
@article{ 10.1109/34.161350, author = {J. Konrad and E. Dubois}, title = {Bayesian Estimation of Motion Vector Fields}, journal ={IEEE Transactions on Pattern Analysis and Machine Intelligence}, volume = {14}, number = {9}, issn = {01628828}, year = {1992}, pages = {910927}, doi = {http://doi.ieeecomputersociety.org/10.1109/34.161350}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, }  
RefWorks Procite/RefMan/Endnote  x  
TY  JOUR JO  IEEE Transactions on Pattern Analysis and Machine Intelligence TI  Bayesian Estimation of Motion Vector Fields IS  9 SN  01628828 SP910 EP927 EPD  910927 A1  J. Konrad, A1  E. Dubois, PY  1992 KW  Bayesian estimation; minimum expected cost estimation; picture processing; 2D motion vector fields; timevarying images; deterministic structural model; vector Markov random fields; piecewise smooth model; maximum a posteriori probability; stochastic relaxation; Gibbs sampler; state space; Bayes methods; estimation theory; Markov processes; picture processing; probability; statespace methods VL  14 JA  IEEE Transactions on Pattern Analysis and Machine Intelligence ER   
A stochastic approach to the estimation of 2D motion vector fields from timevarying images is presented. The formulation involves the specification of a deterministic structural model along with stochastic observation and motion field models. Two motion models are proposed: a globally smooth model based on vector Markov random fields and a piecewise smooth model derived from coupled vectorbinary Markov random fields. Two estimation criteria are studied. In the maximum a posteriori probability (MAP) estimation, the a posteriori probability of motion given data is maximized, whereas in the minimum expected cost (MEC) estimation, the expectation of a certain cost function is minimized. Both algorithms generate sample fields by means of stochastic relaxation implemented via the Gibbs sampler. Two versions are developed: one for a discrete state space and the other for a continuous state space. The MAP estimation is incorporated into a hierarchical environment to deal efficiently with large displacements.
[1] P. Anandan, "Measuring visual motion from image sequences," Ph.D. thesis, Univ. of Mass., May 1987.
[2] M. Bertero, T. A. Poggio, and V. Torre, "Illposed problems in early vision,"Proc. IEEE, vol. 76, pp. 869889, Aug. 1988.
[3] J. Besag, "Spatial interaction and the statistical analysis of lattice systems,"J. Roy. Stat. Soc., vol. B 36, pp. 192236, 1974.
[4] J. Besag, "On the statistical analysis of dirty pictures,"J. Roy. Stat. Soc., vol. B 48, pp. 259279, 1986.
[5] P. Bouthemy and P. Lalande, "Motion detection in an image sequence using Gibbs distributions," inProc. IEEE Int. Conf. Acoust. Speech Signal Processing, May 1989, pp. 16511654.
[6] P. Burt, "Fast filter transforms for image processing,"Comput. Vision Graphics Image Processing, vol. 16, pp. 2051, 1981.
[7] H. Derin and H. Elliott, "Modeling and segmentation of noisy and textured images using Gibbs random fields,"IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI9, pp. 3955, Jan. 1987.
[8] E. Dubois, "The sampling and reconstruction of timevarying imagery with application in video systems,"Proc. IEEE, vol. 73, pp. 502522, Apr. 1985.
[9] W. Enkelmann, "Investigations of multigrid algorithms for the estimation of optical flow fields in image sequences,"Comput. Vision Graphics Image Processing, vol. 43, pp. 150177, 1988.
[10] S. Geman and D. Geman, "Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images,"IEEE Trans. Patt. Anal. Machine Intell., vol. PAMI6, pp. 721741, Nov. 1984.
[11] M. A. Gennert and S. Negahdaripour, "Relaxing the brightness constancy assumption in computing optical flow," Artificial Intell. Lab., Mass. Inst. Technol., MIT AI Memo 975, June 1987.
[12] F. Glazer, "Hierarchical motion detection," Ph.D. thesis, Univ. of Mass., Feb. 1987.
[13] M. Hassner and J. Sklansky, "The use of Markov random fields as models for texture" inImage Modeling(E. A. Rosenfeld, Ed.). New York: Academic, 1981, pp. 185198.
[14] F. Heitz and P. Bouthemy, "Motion estimation and segmentation using a global Bayesian approach," inProc. IEEE Int. Conf. Acoust. Speech Signal Processing, 1990, pp. 23052308.
[15] F. Heitz and P. Bouthemy, "Multimodal motion estimation and segmentation using Markov random fields," inProc. IEEE Int. Conf. Patt. Recogn., June 1990, pp. 378383.
[16] E. Hildreth, "Computation underlying the measurement of visual motion,"Artificial Intell., vol. 23, pp. 309354, 1984.
[17] B. P. Horn and B. Schunck, "Determining optical flow,"Artificial Intell., vol. 17, pp. 185203, 1981.
[18] J. Hutchinson, C. Koch, J. Luo, and C. Mead, "Computing motion using analog and binary resistive networks,"Comput., vol. 21, pp. 5263, Mar. 1988.
[19] R. G. Keys, "Cubic convolution interpolation for digital image processing,"IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP29, pp. 11531160, 1981.
[20] S. Kirkpatrick, C. Gelatt, Jr., and M. Vecchi, "Optimization by simulated annealing,"Sci., vol. 220, pp. 671680, May 1983.
[21] J. Konrad and E. Dubois, "Estimation of image motion fields: Bayesian formulation and stochastic solution," inProc. IEEE Int. Conf. Acoust. Speech Signal Processing, Apr. 1988, pp. 10721075.
[22] J. Konrad and E. Dubois, "Multigrid Bayesian estimation of image motion fields using stochastic relaxation," inProc. IEEE Int. Conf. Comput. Vision, Dec. 1988, pp. 354362.
[23] J. Konrad and E. Dubois, "Motioncompensated interpolation for TV framerate conversion," Tech. Rep. 26, INRSTélécommunications, Oct. 1988.
[24] J. Konrad and E. Dubois, "Bayesian estimation of motion fields from image sequences," Ph.D. thesis, Dept. of Elect. Eng., McGill Univ., Montreal, June 1989.
[25] J. Konrad and E. Dubois, "Bayesian estimation of discontinuous motion in images using simulated annealing," inProc. Conf. Vision Interface VI'89June 1989, pp. 5160.
[26] J. Konrad and E. Dubois, "Comparison of stochastic and deterministic solution methods in Bayesian estimation of 2D motion,"Image Vision Comput., vol. 9, pp. 215228, Aug. 1991.
[27] J. Marroquin, "Probablistic solution of inverse problems," Ph.D. thesis, Mass. Inst. Technol., Cambridge, Sept. 1985.
[28] D. Martinez, "Modelbased motion estimation and its application to restoration and interpolation of motion pictures," Ph.D. thesis, Mass. Inst. Technol., Cambridge, Aug. 1986.
[29] N. Metropolis, A. Rosenbluth, M. Rosenbluth, H. Teller, and E. Teller, "Equation of state calculations by fast computing machines,"J. Chem. Phys., vol. 21, pp. 10871092, June 1953.
[30] D. W. Murray and B. F. Buxton, "Scene segmentation from visual motion using global optimization,"IEEE Trans. Patt. Analy. Machine Intell., vol. PAMI9, pp. 161180, 1987.
[31] H.H. Nagel and W. Enkelmann, "An investigation of smoothness constraints for the estimation of displacement vector fields from image sequences,"IEEE Trans. Patt. Anal. Machine Intell., vol. PAMI8, no. 5, pp. 565593, Sept. 1986.
[32] S. Ross,Applied Probability Models with Optimization Applications. San Francisco: HoldenDay, 1970.
[33] B. Schunck, "Motion segmentation and estimation," Ph.D. thesis, Mass. Inst. Technol., Cambridge, May 1983.
[34] O. Tretiak and L. Pastor, "Velocity estimation from image sequences with second order differential operators," inProc. IEEE Int. Conf. Patt. Recogn., July 1984, pp. 1619.
[35] J. Woods, "Twodimensional discrete Markov random fields,"IEEE Trans. Inform. Theory,vol. IT18, pp. 232240, 1972.