Abstract

Scene analysis is a prior stage in many computer vision and robotics applications. Thanks to recent depth camera, we propose a fast plane segmentation approach for obstacle detection in indoor environments. The proposed method Fast RANdom Sample Consensus (FRANSAC) involves three steps: data input, data preprocessing and 3D RANSAC. Firstly, range data, obtained from 3D camera, is converted into 3D point clouds. Next, a preprocessing stage is introduced where a pass through and voxel grid filters are applied. Finally, planes are estimated using a modified 3D RANSAC. The experimental results demonstrate that our approach can segment planes and detect obstacles about 7 times faster than the standard RANSAC without losing the discriminative power.

References

• Y. M. Kim, N. J. Mitra, D.-M. Yan, and L. Guibas, “Acquiring 3D indoor environments with variability and repetition,” ACM Trans. Graph., vol. 31, no. 6, p. 1, 2012.
• C. V. Nguyen, S. Izadi, and D. Lovell, “Modeling kinect sensor noise for improved 3D reconstruction and tracking,” Proc. - 2nd Jt. 3DIM/3DPVT Conf. 3D Imaging, Model. Process. Vis. Transm. 3DIMPVT 2012, pp. 524–530, 2012.
• M. Niessner, M. Zollhöfer, S. Izadi, and M. Stamminger, “Real-time 3D Reconstruction at Scale Using Voxel Hashing,” ACM Trans. Graph., vol. 32, no. 6, p. 169:1--169:11, 2013.
• M. Zollh et al., “Real-time Non-rigid Reconstruction using an RGB-D Camera,” 2013.
• A. Dai, M. Nießner, M. Zollhöfer, S. Izadi, and C. THEOBALT, “BundleFusion: Real-time Globally Consistent 3D Reconstruction using On-the-fly Surface Re-integration,” arXiv.org, 2016.
• M. Innmann, M. Zollhöfer, M. Nießner, C. Theobalt, and M. Stamminger, “VolumeDeform: Real-time Volumetric Non-rigid Reconstruction,” pp. 1–17, 2016.
• L. Alexandre, “3D Object Recognition using Convolutional Neural Networks with Transfer Learning between Input Channels,” 13th Int. Conf. Intell. Auton. Syst., 2014.
• S. Aigerim, A. Askhat, and A. Yedilkhan, “Recognition of 3D object using Kinect,” Appl. Inf. Commun. Technol. (AICT), 9th Int. Conf., pp. 341–346, 2015.
• G. Pang and U. Neumann, “Fast and Robust Multi-view 3D Object Recognition in Point Clouds,” 3D Vis. (3DV), Int. Conf., pp. 171–179, 2015.
• O. Hilliges, Kim, S. Izadi, M. Weiss, and A. Wilson, “Holodesk: Direct 3D interactions with a situated see-through display,” Proc. CHI 2012, pp. 2421–2430, 2012.
• A. Wilson, H. Benko, S. Izadi, and O. Hilliges, “Steerable augmented reality with the beamatron,” UIST ’12 Proc. 25th Annu. ACM Symp. User interface Softw. Technol., pp. 413–422, 2012.
• R. Du, “Video Fields : Fusing Multiple Surveillance Videos into a Dynamic Virtual Environment,” pp. 1–8, 2016.
• M. a. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM, vol. 24, no. 6, pp. 381–395, 1981.
• P. J. Huber, Robust Statistics, vol. 82, no. 3. 1981.
• C. V. Stewart, “Robust Parameter Estimation in Computer Vision,” SIAM Rev., vol. 41, no. 3, pp. 513–537, 1999.
• J. Matas and O. Chum, “Randomized RANSAC with Td,d test,” Image Vis. Comput., vol. 22, no. 10 SPEC. ISS., pp. 837–842, 2004.
• P. Gyawali and J. McGough, “Simulation of detecting and climbing a ladder for a humanoid robot,” IEEE Int. Conf. Electro Inf. Technol., 2013.
• J. Pardeiro, J. V. Gómez, D. Álvarez, and L. Moreno, “Learning-based floor segmentation and reconstruction,” Adv. Intell. Syst. Comput., vol. 253, pp. 307–320, 2014.
• A. Hidalgo-Paniagua, M. A. Vega-Rodríguez, N. Pavón, and J. Ferruz, “A Comparative Study of Parallel RANSAC Implementations in 3D Space,” Int. J. Parallel Program., vol. 43, no. 5, pp. 703–720, 2014.
• L. Dung, C. Huang, and Y. Wu, “Implementation of RANSAC Algorithm for Feature-Based Image Registration,” J. Comput. Commun., vol. 2013, no. November, pp. 46–50, 2013.
• J. W. Tang, N. Shaikh-Husin, and U. U. Sheikh, “FPGA implementation of RANSAC algorithm for real-time image geometry estimation,” Proceeding - 2013 IEEE Student Conf. Res. Dev. SCOReD 2013, no. December, pp. 290–294, 2015.
• J. Vourvoulakis, J. Lygouras, and J. Kalomiros, “Acceleration of RANSAC algorithm for images with affine transformation,” 2016 IEEE Int. Conf. Imaging Syst. Tech., pp. 60–65, 2016.
• J. Sturm, N. Engelhard, F. Endres, W. Burgard, and D. Cremers, “A benchmark for the evaluation of RGB-D SLAM systems,” in IEEE International Conference on Intelligent Robots and Systems, 2012, pp. 573–580.