Interactive System for Objects Recognition using 3D Camera, Point Cloud Segmentation and Augmented Reality User Interface

Matej Nikorovic, Radoslav Gargalik, Zoltan Tomori

Abstract

Depth-sensing cameras are frequently used in computer vision and augmented reality applications as a key component of the point cloud acquisition system as well as a natural user interface tool. We integrated both these functions into the automatic objects recognition system based on the machine learning. Acquired point cloud is segmented by the region growing algorithm exploiting smoothness constraint as a homogeneity criterion. Segmented objects lying on the ground plane are recognized via the supervised machine learning and the corresponding label is is projected near to the object. Natural user interface controls the learning process as well as the mode of operation. System is proper for specific environments like e.g. science center (museum).

References

  1. Arthur, D. and Vassilvitskii, S. (2007). k-means++: The advantages of careful seeding. In SODA 7807 Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms.
  2. As'ari, M., Sheikh, U., and Supriyanto, E. (2013). 3d shape descriptor for object recognition based on kinect-like depth image. In Image and Vision Computing. ScienceDirect.
  3. Boykov, Y. and Kolmogorov, V. (2004). An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision. In Pattern Analysis and Machine Intelligence, IEEE Transactions.
  4. Castillo, E., Liang, J., and Zhao, H. (2013). Point cloud segmentation and denoising via constrained nonlinear least squares normal estimates. In Innovations for Shape Analysis. Springer Berlin Heidelberg.
  5. Dal Mutto, C., Zanuttigh, P., and Cortelazzo, G. M. (2010). Scene segmentation by color and depth information and its applications.
  6. Dupuis, J., Paulus, S., Behmann, J., Plumer, L., and Kuhlmann, H. (2014). A multi-resolution approach for an automated fusion of different low-cost 3d sensors. In Sensors 2014.
  7. Golovinskiy, A. and Funkhouser, T. (2009). Min-cut based segmentation of point clouds. In IEEE Workshop on Search in 3D and Video (S3DV) at ICCV.
  8. Klasing, K., Althoff, D., Wollherr, D., and Buss, M. (2009). Comparison of surface normal estimation methods for range sensing applications. In Robotics and Automation. IEEE.
  9. Mine, M., Rose, D., Yang, B., Vanbaar, J., and Grundhofer, A. (2012). Projection-based augmented reality in disney theme parks. In Computer 45, 7, pages 32-40.
  10. Rabbani, T., van den Heuvel, F. A., and Vosselman, G. (2006). Segmentation of point clouds using smoothness constraint. In Robotics and Automation. ISPRS Commission V Symposium 'Image Engineering and Vision Metrology'.
  11. Sedlacek, D. and Zara, J. (2009). Graph cut based pointcloud segmentation for polygonal reconstruction. In Advances in Visual Computing.
  12. Tomori, Z., Vanko, P., and Vaitovic, B. (2015). Using of low-cost 3d cameras to control interactive exhibits in science center. In Sincak, P., Hartono, P., Vircikova, M., Vascak, J., and Jaksa, R. (Eds.): 'Emergent Trends in Robotics and Intelligent Systems: Where is the role of intelligent technologies in the next generation of robots? Springer.
  13. Vosselman, G. (2013). Point cloud segmentation for urban scene classification. In International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. ISPRS2013-SSG.
  14. Wang, J. and Shan, J. (2009). Segmentation of lidar point clouds for building extraction. ASPRS 2009 Annual Conference.
  15. Zhan, Q., Liang, Y., and Xiao, Y. (2009). Color-based segmentation of point clouds. In Laser scanning 2009. IAPRS.
Download


Paper Citation


in Harvard Style

Nikorovic M., Gargalik R. and Tomori Z. (2015). Interactive System for Objects Recognition using 3D Camera, Point Cloud Segmentation and Augmented Reality User Interface . In Doctoral Consortium - DCVISIGRAPP, (VISIGRAPP 2015) ISBN , pages 31-36


in Bibtex Style

@conference{dcvisigrapp15,
author={Matej Nikorovic and Radoslav Gargalik and Zoltan Tomori},
title={Interactive System for Objects Recognition using 3D Camera, Point Cloud Segmentation and Augmented Reality User Interface},
booktitle={Doctoral Consortium - DCVISIGRAPP, (VISIGRAPP 2015)},
year={2015},
pages={31-36},
publisher={SciTePress},
organization={INSTICC},
doi={},
isbn={},
}


in EndNote Style

TY - CONF
JO - Doctoral Consortium - DCVISIGRAPP, (VISIGRAPP 2015)
TI - Interactive System for Objects Recognition using 3D Camera, Point Cloud Segmentation and Augmented Reality User Interface
SN -
AU - Nikorovic M.
AU - Gargalik R.
AU - Tomori Z.
PY - 2015
SP - 31
EP - 36
DO -