Abstract
In this paper, we present a patch-based direct visual odometry (DVO) that is robust to illumination changes at a sequence of stereo images. Illumination change violates the photo-consistency assumption and degrades the performance of DVO, thus, it should be carefully handled during minimizing the photometric error. Our approach divides an incoming image into several buckets, and patches inside each bucket own its unique affine illumination parameter to account for local illumination changes for which the global affine model fails to account, then it aligns small patches placed at temporal images. We do not distribute affine parameters to each patch since this yields huge computational load. Furthermore, we propose a prior weight as a function of the previous pose in a constant velocity model which implies that the faster a camera moves, the more likely it maintains the constant velocity model. Lastly, we verify that the proposed algorithm outperforms the global affine illumination model at the publicly available micro aerial vehicle and the planetary rover dataset which exhibit irregular and partial illumination changes due to the automatic exposure of the camera and the strong outdoor sunlight, respectively.
Similar content being viewed by others
References
S. Davide and F. Fraundorfer, “Visual odometry [tutorial],” IEEE Robotics & Automation Magazine, vol. 18, no. 4, pp. 80–92, December 2011.
C. Yang, M Maimone, and L. Matthies, “Visual odometry on the Mars exploration rovers,” Proc. of IEEE International Conference on Systems, Man and Cybernetics, pp. 903–910, October 2005.
C. Forster, M. Pizzoli, and D. Scaramuzza, “SVO: fast semi-direct monocular visual odometry,” Proc. of IEEE International Conference on Robotics & Automation, pp. 15–22, May 2014.
D. Nister, O. Naroditsky, and J. Bergen, “Visual odometry,” Proc. of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, July 2004.
F. Steinbrucker, J. Sturm, and D. Cremers, “Real-time visual odometry from dense RGB-D images,” Proc. of IEEE International Conference on Computer Vision Workshops, pp. 719–722, November 2011.
V. Q. Nhat and G. Lee, “Illumination invariant object tracking with adaptive sparse representation,” International Journal of Control, Automation and Systems, vol. 12, no. 1, pp. 195–201, February 2014.
P. Kim, B. Coltin, O. Alexandrov, and H. J. Kim, “Robust visual localization in changing lighting conditions,” Proc. of IEEE International Conference on Robotics and Automation, pp. 5447–5252, June 2017.
S. Klose, P. Heise, and A. Knoll, “Efficient compositional approaches for real-time robust direct visual odometry from RGB-D data,” Proc. of IEEE International Conference on Intelligent Robots and Systems, pp. 1100–1106, November 2013.
J. Jordan and A. Zell, “Ground plane based visual odometry for RGBD-cameras using orthogonal projection,” International Federation of Automatic Control on Intelligent Autonomous Vehicles, pp. 108–113, June 2016.
H. Jin, P. Favaro, and S. Soatto, “Real-time feature tracking and outlier rejection with changes in illumination,” Proc. of IEEE International Conference on Computer Vision, pp. 684–689, July 2001.
J. Engel, J. Stuckler, and D. Cremers, “Large-scale direct SLAM with stereo cameras,” Proc. of IEEE International Conference on Intelligent Robots and Systems, pp. 1935–1942, September 2015.
P. Kim, H. Lim, and H. J. Kim, “Robust visual odometry to irregular illumination changes with RGB-D camera,” Proc. of IEEE International Conference on Intelligent Robots and Systems, pp. 3688–3694, October 2015.
B. Kitt, A. Geiger, and H. Lategahn, “Visual odometry based on stereo image sequences with ransac-based outlier rejection scheme,” IEEE Intelligent Vehicles Symposium, pp. 486–492, June 2010.
P. Furgale, P. Carle, J. Enright, and T. D. Barfoot, “The Devon island rover navigation dataset,” The International Journal of Robotics Research, vol. 31, no. 6, pp. 707–713, April 2012.
M. Burri, J. Nikolic, P. Gohl, T. Schneider, J. Rehder, S. Omari, M. W. Achtelik, and R. Siegwart, “The EuRoC micro aerial vehicle datasets,” The International Journal of Robotics Research, vol. 35, no. 10, pp. 1157–1163, April 2016.
J. Engel, J. Sturm, and D. Cremers, “Semi-dense visual odometry for a monocular camera,” IEEE International Conference on Computer Vision, pp. 1449–1456, December 2013.
C. Kerl, J. Sturm, and D. Cremers, “Robust odometry estimation for RGB-D cameras,” Proc. of IEEE International Conference on Intelligent Robots and Systems, pp. 3748–3754, May 2013.
C. Kerl, Odometry from RGB-D Cameras for Autonomous Quadcopters, Master’s Thesis, Technical University of Munich, 2012.
J. Shi, “Good features to track,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 593–600, June 1994.
B. D. Lucas and T. Kanade, “An iterative image registration technique with an application to stereo vision,” Proceedings DARPA Image UnderstandingWorkshop, pp. 121–130, April 1981.
J. Sturm, N. Engelhard, F. Endres,W. Burgard, and D. Cremers “A benchmark for the evaluation of RGB-D SLAM systems,” Proc. of IEEE International Conference on Intelligent Robots and Systems, pp. 573–580, October 2012.
X. Zheng, Z. Moratto, M. Li, and A. I. Mourikis, “Photometric patch-based visual-inertial odometry,” Proc. of IEEE International Conference on Robotics and Automation, pp. 3264–3271, May 2017.
Author information
Authors and Affiliations
Corresponding author
Additional information
Recommended by Associate Editor Kang-Hyun Jo under the direction of Editor Euntai Kim. This work was supported by the Ministry of Science and ICT of the Republic of Korea through the Space Core Technology Development Program under Project NRF-2018M1A3A3A02065722.
Jae Hyung Jung is an M.S. student in the Department of Mechanical and Aerospace Engineering of Seoul National University, Korea. He received the B.S. degree in the Department of Aerospace Engineering from Pusan National University, Korea in 2017. His research interests include visual odometry and vision-aided inertial navigation for mobile robots.
Sejong Heo is a Ph.D. student in the Department of Mechanical and Aerospace Engineering of Seoul National University, Korea. He received the B.S. and M.S. degrees in the Department of Mechanical and Aerospace Engineering from Seoul National University, Korea, in 2008 and 2010, respectively. He worked for Doosan DST in Korea, which is the maker of the high precision INS. His current research topics include the high precision inertial navigation, Bayesian filtering, nonlinear optimization and vision-aided inertial navigation for land vehicles and mobile robots.
Chan Gook Park received the B.S., M.S., and Ph.D. in control and instrumentation engineering from Seoul National University, South, Korea, in 1985, 1987, and 1993, respectively. He worked with Prof. Jason L. Speyer on peak seeking control for formation flight at the University of California, Los Angeles (UCLA) as a postdoctoral fellow in 1998. From 1994 to 2003, he was with Kwangwoon University, Seoul, Korea, as an associate professor. In 2003, he joined the faculty of the School of Mechanical and Aerospace Engineering at Seoul National University, Korea, where he is currently a professor. From 2009 to 2010, he was a visiting scholar with the Department of Aerospace Engineering at Georgia Institute of Technology, Atlanta, GA. He served as a chair of IEEE AES Korea Chapter until 2009. His current research topics include advanced filtering techniques, high precision INS, GPS/INS integration, MEMSbased pedestrian dead reckoning, and visual inertial navigation.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Jung, J.H., Heo, S. & Park, C.G. Patch-based Stereo Direct Visual Odometry Robust to Illumination Changes. Int. J. Control Autom. Syst. 17, 743–751 (2019). https://doi.org/10.1007/s12555-018-0199-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12555-018-0199-2