Abstract
This paper presents the design and implementation of a new comparative analytical framework for studying the usability of modern high breakdown robust estimators. The emphasis is on finding the intrinsic limits, in terms of size and relative spatial accuracy, of such techniques in solving the emerging challenges of the segmentation of fine structures. A minimum threshold for the distance between separable structures is shown to depend mainly on the scale estimation error. A scale invariant performance measure is introduced to quantify the finite sample bias of the scale estimate of a robust estimator and the measure is evaluated for some state-of-the-art high breakdown robust estimators using datasets containing at least two close but distinct structures with varying distances and inlier ratios. The results show that the new generation of density-based robust estimators (such as pbM-estimator and TSSE) have a poorer performance in problems with datasets containing only a small number of samples in each structure compared with ones based on direct processing of the residuals (such as MSSE). An important message of this paper is that an estimator that performs best in some circumstances, may not be competitive in others: particularly performance on data structures that are relatively large and/or well-separated vs closely spaced fine structures.
Similar content being viewed by others
References
Bab-Hadiashar, A., Suter, D.: Robust optic flow computation. Int. J. Comput. Vis. 29(1), 59–77 (1998)
Qian, G., Chellappa, R., Zheng, Q.: Bayesian algorithms for simultaneous structure from motion estimation of multiple independently moving objects. IEEE Trans. Image Process. 14(1), 94–109 (2005)
Vidal, R., Ma, Y., Soatto, S., Sastry, S.: Two-view multibody structure from motion. Int. J. Comput. Vis. 68(1), 7–25 (2006)
Bab-Hadiashar, A., Gheissari, N.: Range image segmentation using surface selection criterion. IEEE Trans. Image Process. 15(7), 2006–2018 (2006)
Bab-Hadiashar, A., Suter, D.: Range and motion segmentation. In: Bab-Hadiashar, A., Suter, D. (eds.) Data Segmentation and Model Selection for Computer Vision, pp. 119–142. Springer, New York (2000). Chap. 5
Meer, P.: Robust techniques for computer vision. In: Medioni, G., Kang, S. (eds.) Emerging Topics in Computer Vision, pp. 107–190. Prentice Hall, New York (2004). Chap. 3
Chen, H., Meer, P.: Robust regression with projection based M-estimators. In: Proceedings of the Ninth IEEE International Conference on Computer Vision (ICCV’03), pp. 878–885, Nice, France (2003)
Subbarao, R., Meer, P.: Heteroscedastic projection based M-estimators. In: Workshop on Empirical Evaluation Methods in Computer Vision (in Conjunction with CVPR’05), pp. 38–44, San Diego, CA (2005)
Subbarao, R., Meer, P.: Beyond RANSAC: user independent robust regression. In: Workshop on 25 Years of RANSAC (in Conjunction with CVPR’06), pp. 101–108, New York, NY (2006)
Romanesque Churches of the Bourbonnais (France): Image and range data provided by Columbia University, Department of Art, History and Archeology. http://www.mcah.columbia.edu/bourbonnais
Stewart, C.: Bias in robust estimation caused by discontinuities and multiple structures. IEEE Trans. Pattern Anal. Mach. Intell. 19(8), 818–833 (1997)
Hoseinnezhad, R., Bab-Hadiashar, A.: Consistency of robust estimators in multi-structural visual data segmentation. Pattern Recognit. 40, 3677–3690 (2007)
Yu, X., Bui, T., Krzyzak, A.: Robust estimation for range image segmentation and reconstruction. IEEE Trans. Pattern Anal. Mach. Intell. 16(5), 530–538 (1994)
Bab-Hadiashar, A., Suter, D.: Robust segmentation of visual data using ranked unbiased scale estimator. Robotica 17, 649–660 (1999)
Lee, K., Meer, P., Park, R.: Robust adaptive segmentation of range images. IEEE Trans. Pattern Anal. Mach. Intell. 20(2), 200–205 (1998)
Wang, H., Suter, D.: Robust adaptive-scale parametric model estimation for computer vision. IEEE Trans. Pattern Anal. Mach. Intell. 26(11), 1459–1474 (2004)
Rousseeuw, P., Leroy, A.: Robust Regression and Outlier Detection, pp. 1–20. Wiley, New York (2003). Chap. 1
Huber, P.: Robust Statistics. Wiley, New York (2004). Chap. 5
Li, G.: Robust regression. In: Hoaglin, D., Mosteller, F., Tukey, J. (eds.) Exploring Data Tables, Trends and Shapes, pp. 281–343. Wiley, New York (1985)
Bab-Hadiashar, A., Suter, D.: Motion segmentation: a robust approach. In: Proceedings of IEEE International Workshop on the Interpretation of Visual Motion (IVM 98), pp. 3–9, Santa Barbara, California (1998)
Cheng, Y.: Mean shift, mode seeking, and clustering. IEEE Trans. Pattern Anal. Mach. Intell. 17(8), 790–799 (1995)
Comaniciu, D., Meer, P.: Mean shift: a robust approach toward feature space analysis. IEEE Trans. Pattern Anal. Mach. Intell. 24(5), 603–619 (2002)
Fashing, M., Tomasi, C.: Mean shift is a bound optimization. IEEE Trans. Pattern Anal. Mach. Intell. 27(3), 471–474 (2005)
Fischler, M., Bolles, R.: Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. Commun. Assoc. Comput. Mach. 24(6), 381–395 (1981)
Gander, W., Gautschi, W.: Adaptive quadrature—revisited. BIT Numer. Math. 40, 84–101 (2000)
Subbarao, R., Meer, P.: Subspace estimation using projection based M-estimators over Grassman manifolds. In: Ales Leonardis, A.P. (ed.) 9th European Conference on Computer Vision–ECCV’06, Graz, Austria, May 2006, pp. 301–312. Springer, Berlin (2006)
Ronchetti, E.: Robust regression methods and model selection. In: Bab-Hadiashar, A., Suter, D. (eds.) Data Segmentation and Model Selection for Computer Vision, pp. 31–40. Springer, New York (2000). Chap. 2
Author information
Authors and Affiliations
Corresponding author
Additional information
Reza Hoseinnezhad was with Swinburne University of Technology when the methodologies and results reported in this paper were developed.
Rights and permissions
About this article
Cite this article
Hoseinnezhad, R., Bab-Hadiashar, A. & Suter, D. Finite Sample Bias of Robust Estimators in Segmentation of Closely Spaced Structures: A Comparative Study. J Math Imaging Vis 37, 66–84 (2010). https://doi.org/10.1007/s10851-010-0193-7
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10851-010-0193-7