Fast Incremental Objects Identification and Localization using Cross-correlation on a 6 DoF Voting Scheme

Mauro Antonello, Alberto Pretto, Emanuele Menegatti

2014

Abstract

In this work, we propose a sparse features-based object recognition and localization system, well suited for online learning of new objects. Our method takes advantages of both depth and ego-motion information, along with salient feature descriptors information, in order to learn and recognize objects with a scalable approach. We extend the conventional probabilistic voting scheme for object the recognition task, proposing a correlation-based approach in which each object-related point feature contributes in a 6-dimensional voting space (i.e., the 6 degrees-of-freedom, DoF, object position) with a continuous probability density distribution (PDF) represented by a Mixture of Gaussian (MoG). A global PDF is then obtained adding the contribution of each feature. The object instance and pose are hence inferred exploiting an efficient mode-finding method for mixtures of Gaussian distributions. The special properties of the convolution operator for the MoG distributions, combined with the sparsity of the exploited data, provide our method with good computational efficiency and limited memory requirements, enabling real-time performances also in robots with limited resources.

References

  1. Carreira-Perpinan, M. (2000). Mode-finding for mixtures of Gaussian distributions. Pattern Anal. Mach. Learn., pages 1-23.
  2. Declercq, A. and Piater, J. H. (2008). Online learning of gaussian mixture models: a two-level approach.
  3. Dempster, A., Laird, N., and Rubin, D. (1977). Maximum likelihood from incomplete data via the EM algorithm. J. R. Stat. Soc., 39(1):1-38.
  4. Hall, P., Hicks, Y., and Robinson, T. (2005). A method to add Gaussian mixture models.
  5. Lai, K., Bo, L., Ren, X., and Fox, D. (2011a). A large-scale hierarchical multi-view RGB-D object dataset. ICRA, pages 1817-1824.
  6. Lai, K., Bo, L., Ren, X., and Fox, D. (2011b). Sparse distance learning for object recognition combining RGB and depth information. ICRA, (1):4007-4013.
  7. Marvel, J. a., Hong, T.-H., and Messina, E. (2012). 2011 Solutions in Perception Challenge Performance Metrics and Results. Proc. Work. Perform. Metrics Intell. Syst. - Permis 7812, page 59.
  8. Ognjen, A. and Cipolla, R. (2005). Incremental Learning of Temporally-Coherent Gaussian Mixture Models.
  9. Rusu, R. and Bradski, G. (2010). Fast 3d recognition and pose using the viewpoint feature histogram. Intell. Robot. . . . , pages 2155-2162.
  10. Rusu, R. B. (2009). Fast Point Feature Histograms (FPFH) for 3D registration. Robot. Autom. 2009. . . . , pages 3212-3217.
  11. Song, M. and Wang, H. (2005). Highly efficient incremental estimation of Gaussian mixture models for online data stream clustering. pages 174-183.
  12. Tang, J. and Miller, S. (2012). A textured object recognition pipeline for color and depth image data. Robot. Autom.
  13. Tsotsos, K., Pretto, A., and Soatto, S. (2012). VisualInertial Ego-Motion Estimation for Humanoid Platforms. In IEEE-RAS Int. Conf. Humanoid Robot., pages 704-711.
  14. Vaskevicius, N. and Pathak, K. (2012). The jacobs robotics approach to object recognition and localization in the context of the ICRA'11 Solutions in Perception Challenge. Robot. Autom. . . . , pages 3475-3481.
  15. Wohlkinger, W., Aldoma, A., Rusu, R. B., and Vincze, M. (2012). 3DNet: Large-scale object class recognition from CAD models. ICRA, pages 5384-5391.
  16. Xie, Z., Singh, A., Uang, J., Narayan, K., and Abbeel, P. (2013). Multimodal Blending for High-Accuracy Instance Recognition. IROS.
Download


Paper Citation


in Harvard Style

Antonello M., Pretto A. and Menegatti E. (2014). Fast Incremental Objects Identification and Localization using Cross-correlation on a 6 DoF Voting Scheme . In Proceedings of the 9th International Conference on Computer Graphics Theory and Applications - Volume 1: WARV, (VISIGRAPP 2014) ISBN 978-989-758-002-4, pages 499-504. DOI: 10.5220/0004873604990504


in Bibtex Style

@conference{warv14,
author={Mauro Antonello and Alberto Pretto and Emanuele Menegatti},
title={Fast Incremental Objects Identification and Localization using Cross-correlation on a 6 DoF Voting Scheme},
booktitle={Proceedings of the 9th International Conference on Computer Graphics Theory and Applications - Volume 1: WARV, (VISIGRAPP 2014)},
year={2014},
pages={499-504},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0004873604990504},
isbn={978-989-758-002-4},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 9th International Conference on Computer Graphics Theory and Applications - Volume 1: WARV, (VISIGRAPP 2014)
TI - Fast Incremental Objects Identification and Localization using Cross-correlation on a 6 DoF Voting Scheme
SN - 978-989-758-002-4
AU - Antonello M.
AU - Pretto A.
AU - Menegatti E.
PY - 2014
SP - 499
EP - 504
DO - 10.5220/0004873604990504