Batch Gating for Data Association in Monocular SLAM

Abstract:

This work describes the development and implementation of a single-camera SLAM system, introducing a novel data validation algorithm. A 6-DOF monocular SLAM method developed is based on the Delayed Inverse-Depth (DI-D) Feature Initialization, with the addition of a new data association batch validation technique, the Highest Order Hypothesis Compatibility Test, HOHCT. The DI-D initializes new features in the system defining single hypothesis for the initial depth of features by stochastic triangulation. The HOHCT is based on evaluation of statistically compatible hypotheses, and search algorithm designed to exploit the Delayed Inverse-Depth technique characteristics. Experiments with real data are presented in order to validate the performance of the system.

• Periodical: Applied Mechanics and Materials (Volume 319)
• Pages: 295-301
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.319.295
• Online since: May 2013
• Authors: Edmundo Guerra, Rodrigo Munguia, Yolanda Bolea, Antoni Grau
• Keywords: Computer Science, Monocular SLAM, Only-Bearing Sensor, Robotics

References

[1] Chen-Han H, Chieh-Chih W (2011). Achieving undelayed initialization in monocular SLAM with generalized objects using velocity estimate-based classification. ICRA 2011 Proceedings.

DOI: 10.1109/icra.2011.5979786

[2] Jin H, Favaro P, Soatto S (2003). A Semi-Direct Approach to Structure from Motion. The Visual Computer. v. 19(6), pp.377-394.

DOI: 10.1007/s00371-003-0202-6

[3] Montiel JMM, Civera J, Davison A (2006). Unified inverse depth parameterization for monocular SLAM. Robotics: Science and Systems Conference.

DOI: 10.15607/rss.2006.ii.011

[4] Williams B, Reid I (2010). On combining visual SLAM and visual odometry. Preoceedings of the International Conference on Robotics and Automation, pp.3494-3500, 3-7 May (2010)

DOI: 10.1109/robot.2010.5509248

[5] Strelow S, Sanjiv D (2003). Online motion estimation from image and inertial measurements. Workshop on Integration of Vision and Inertial Sensors INERVIS.

DOI: 10.1109/cvprw.2003.10073

[6] Kwok NM, Dissanayake G (2005). Bearing-only SLAM using a SPRT based gaussian sum filter. IEEE International Conference on Robotics and Automation.

DOI: 10.1109/robot.2005.1570264

[7] Davison A, Gonzalez Y, Kita N (2004). Real-Time 3D SLAM with wide-angle vision. IFAC Symposium on Intelligent Autonomous Vehicles.

DOI: 10.1016/s1474-6670(17)32089-x

[8] Neira J, Tardos JD (2001). Data association in stochastic mapping using the joint compatibility test. IEEE Transaction on Robotics and Automation, vol.17(6), p.890–897.

DOI: 10.1109/70.976019

[9] Bailey T (2002). Mobile robot localisation and mapping in extensive outdoor environments. Ph.D. dissertation, Univ. Sydney, Australian Ctr. Field Robotics.

[10] Civera J, Grasa OG, Davison AJ, Montiel JMM (2009). 1-point RANSAC for EKF-based Structure from Motion. International Conference on Intelligent Robots and Systems.

DOI: 10.1109/iros.2009.5354410

[11] Munguia R, Grau A. (2012) Monocular SLAM for Visual Odometry: a full Approach to the Delayed Inverse-Depth Feature Initialization Method. Mathematical Problems in Engineering. Avaliable: http://www.hindawi.com/journals/mpe/aip/676385. Accessed 2012 August 2.

DOI: 10.1155/2012/676385

[12] Munguia R, Grau A (2008). Delayed Inverse Depth Monocular SLAM. 17th IFAC World Congress.

DOI: 10.3182/20080706-5-kr-1001.00399

[13] Chatterjee C, Roychowdhury V (2000). Algorithms for coplanar camera calibration. Machine Vision and Applications, vol. 12, pp.84-97.

DOI: 10.1007/s001380050127

[14] Davison A, Murray D (1998). Mobile Robot Localisation using Active Vision. Proceedings of the European Conference on Computer Vision.