EVSAC: Accelerating Hypotheses Generation by Modeling Matching Scores with Extreme Value Theory

Algorithms based on RANSAC that estimate models using feature correspondences between images can slow down tremendously when the percentage of correct correspondences (inliers) is small. In this paper, we present a probabilistic parametric model that allows us to assign confidence values for each matching correspondence and therefore accelerates the generation of hypothesis models for RANSAC under these conditions. Our framework leverages Extreme Value Theory to accurately model the statistics of matching scores produced by a nearest-neighbor feature matcher. Using a new algorithm based on this model, we are able to estimate accurate hypotheses with RANSAC at low inlier ratios significantly faster than previous state-of-the-art approaches, while still performing comparably when the number of inliers is large. We present results of homography and fundamental matrix estimation experiments for both SIFT and SURF matches that demonstrate that our method leads to accurate and fast model estimations.

[1]  Cordelia Schmid,et al.  A Comparison of Affine Region Detectors , 2005, International Journal of Computer Vision.

[2]  Sudeep Sarkar,et al.  BLOGS: Balanced local and global search for non-degenerate two view epipolar geometry , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[3]  Luc Van Gool,et al.  Combined Depth and Outlier Estimation in Multi-View Stereo , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[4]  Jan-Michael Frahm,et al.  A Comparative Analysis of RANSAC Techniques Leading to Adaptive Real-Time Random Sample Consensus , 2008, ECCV.

[5]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[6]  E. Castillo Extreme value and related models with applications in engineering and science , 2005 .

[7]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[8]  Jiri Matas,et al.  Matching with PROSAC - progressive sample consensus , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[9]  Frank Dellaert,et al.  GroupSAC: Efficient consensus in the presence of groupings , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[10]  Andrew Zisserman,et al.  MLESAC: A New Robust Estimator with Application to Estimating Image Geometry , 2000, Comput. Vis. Image Underst..

[11]  Jiri Matas,et al.  Locally Optimized RANSAC , 2003, DAGM-Symposium.

[12]  Torsten Sattler,et al.  SCRAMSAC: Improving RANSAC's efficiency with a spatial consistency filter , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[13]  Ali S. Hadi,et al.  Extreme Value and Related Models with Applications in Engineering and Science , 2004 .

[14]  David W. Murray,et al.  Guided Sampling and Consensus for Motion Estimation , 2002, ECCV.

[15]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[16]  Bernhard P. Wrobel,et al.  Multiple View Geometry in Computer Vision , 2001 .

[17]  Matthew Turk,et al.  SWIGS: A Swift Guided Sampling Method , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[18]  Ilan Shimshoni,et al.  Balanced Exploration and Exploitation Model Search for Efficient Epipolar Geometry Estimation , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[19]  Anderson Rocha,et al.  Meta-Recognition: The Theory and Practice of Recognition Score Analysis , 2011, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[20]  Jan-Michael Frahm,et al.  RECON: Scale-adaptive robust estimation via Residual Consensus , 2011, 2011 International Conference on Computer Vision.

[21]  Tat-Jun Chin,et al.  Accelerated Hypothesis Generation for Multistructure Data via Preference Analysis , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.