Subscribe
Issue No.08 - August (2008 vol.30)
pp: 1472-1482
ABSTRACT
A randomized model verification strategy for RANSAC is presented. The proposed method finds, like RANSAC, a solution that is optimal with user-specified probability. The solution is found in time that is (i) close to the shortest possible and (ii) superior to any deterministic verification strategy. A provably fastest model verification strategy is designed for the (theoretical) situation when the contamination of data by outliers is known. In this case, the algorithm is the fastest possible (on average) of all randomized \\RANSAC algorithms guaranteeing a confidence in the solution. The derivation of the optimality property is based on Wald's theory of sequential decision making, in particular a modified sequential probability ratio test (SPRT). Next, the R-RANSAC with SPRT algorithm is introduced. The algorithm removes the requirement for a priori knowledge of the fraction of outliers and estimates the quantity online. We show experimentally that on standard test data the method has performance close to the theoretically optimal and is 2 to 10 times faster than standard RANSAC and is up to 4 times faster than previously published methods.
INDEX TERMS
G.3.m Robust regression, I.4.1.b Imaging geometry
CITATION
Ondřej Chum, Jiří Matas, "Optimal Randomized RANSAC", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol.30, no. 8, pp. 1472-1482, August 2008, doi:10.1109/TPAMI.2007.70787
REFERENCES
 [1] Proc. IEEE Int'l Workshop “25 Years of RANSAC” in conjunction with CVPR '06 (RANSAC25 '06), http://cmp.felk.cvut.czcvpr06-ransac, 2006. [2] D. Capel, “An Effective Bail-Out Test for RANSAC Consensus Scoring,” Proc. British Machine Vision Conf. (BMVC '05), pp. 629-638, 2005. [3] O. Chum and J. Matas, “Randomized RANSAC with T(d, d) Test,” Proc. British Machine Vision Conf. (BMVC '02), vol. 2, BMVA, pp.448-457, 2002. [4] O. Chum, J. Matas, and J. Kittler, “Locally Optimized RANSAC,” Proc. Ann. Symp. German Assoc. for Pattern Recognition (DAGM '03), 2003. [5] O. Chum, T. Werner, and J. Matas, “Epipolar Geometry Estimation via RANSAC Benefits from the Oriented Epipolar Constraint,” Proc. Int'l Conf. Pattern Recognition (ICPR '04), vol. 1, pp.112-115, Aug. 2004. [6] M. Fischler and R. Bolles, “Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography,” Comm. ACM, vol. 24, no. 6, pp. 381-395, June 1981. [7] R. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision, second ed. Cambridge Univ., 2003. [8] P.M. Lee, Sequential Probability Ratio Test. Univ. of York, www.york.ac.uk/depts/maths/teaching/pml/ aissprt.ps, 2007. [9] J. Matas and O. Chum, “Randomized RANSAC with ${T}_{d, d}$ Test,” Image and Vision Computing, vol. 22, no. 10, pp. 837-842, Sept. 2004. [10] J. Matas and O. Chum, “Randomized RANSAC with Sequential Probability Ratio Test,” Proc. Int'l Conf. Computer Vision, vol. 2, pp.1727-1732, Oct. 2005. [11] D. Myatt, P. Torr, S. Nasuto, J. Bishop, and R. Craddock, “NAPSAC: High Noise, High Dimensional Robust Estimation—It's in the Bag,” Proc. British Machine Vision Conf. (BMVC '02), vol. 2, pp. 458-467, 2002. [12] D. Nister, “Preemptive RANSAC for Live Structure and Motion Estimation,” Proc. Int'l Conf. Computer Vision (ICCV '03), vol. 1, pp.199-206, Oct. 2003. [13] B. Tordoff and D. Murray, “Guided Sampling and Consensus for Motion Estimation,” Proc. Seventh European Conf. Computer Vision (ECCV '02), vol. 1, pp. 82-96, 2002. [14] P. Torr, A. Zisserman, and S. Maybank, “Robust Detection of Degenerate Configurations while Estimating the Fundamental Matrix,” Computer Vision and Image Understanding, vol. 71, no. 3, pp. 312-333, Sept. 1998. [15] P. Torr and A. Zisserman, “MLESAC: A New Robust Estimator with Application to Estimating Image Geometry,” Computer Vision and Image Understanding, vol. 78, pp. 138-156, 2000. [16] A. Wald, Sequential Analysis. Dover, 1947.