# American Institute of Mathematical Sciences

doi: 10.3934/jimo.2021059

## Stereo visual odometry based on dynamic and static features division

 College of Missile Engineering, Rocket Force University of Engineering, Xi'an, Shaanxi 710025, China

* Corresponding author: Guangbin Cai

Received  June 2020 Revised  December 2020 Early access  March 2021

Fund Project: The first author is mainly supported by NSSF of China under Grant (No. 61773387)

Accurate camera pose estimation in dynamic scenes is an important challenge for visual simultaneous localization and mapping, and it is critical to reduce the effects of moving objects on pose estimation. To tackle this problem, a robust visual odometry approach in dynamic scenes is proposed, which can precisely distinguish between dynamic and static features. The key to the proposed method is combining the scene flow and the static features relative spatial distance invariance principle. Moreover, a new threshold is proposed to distinguish dynamic features.Then the dynamic features are eliminated after matching with the virtual map points. In addition, a new similarity calculation function is proposed to improve the performance of loop-closure detection. Finally, the camera pose is optimized after obtaining a closed loop. Experiments have been conducted on TUM datasets and actual scenes, which shows that the proposed method reduces tracking errors significantly and estimates the camera pose precisely in dynamic scenes.

Citation: Hui Xu, Guangbin Cai, Xiaogang Yang, Erliang Yao, Xiaofeng Li. Stereo visual odometry based on dynamic and static features division. Journal of Industrial & Management Optimization, doi: 10.3934/jimo.2021059
##### References:

show all references

##### References:
Stereo camera model
Generation of a visual vocabulary tree
Overview of the proposed algorithm in dynamic scenes
]">Figure 4.  Classification of the scene flow based on angles [26]
Invariance of the relative spatial distance of the static points
Construction of the virtual map points
Three static features selected by the algorithm
Dynamic features obtained by the algorithm
Experiment scene sets
Experimental results of ORB-VO in lab scenes
Experimental results of the proposed method in lab scenes
Loop-closure detection result of the inverse proportional function
Loop-closure detection result of the negative exponential power function
Loop-closure detection result of the negative exponential power function
Comparisons between estimated trajectories and the ground truth in walking sequences
Comparisons between estimated trajectories and the ground truth in sitting sequences
Translation drift and rotational drift of VO method on TUM dataset
 Sequences RMSE of translational drift [m/s] RMSE of rotational drift [$^{\circ}$/s] DVO BaMVO SPW-VO Our Method DVO BaMVO SPW-VO Our Method sitting-static 0.0157 0.0248 0.0231 0.0112 0.6084 0.6977 0.7228 0.3356 sitting-xyz 0.0453 0.0482 0.0219 0.0132 1.4980 1.3885 0.8466 0.5753 sitting-rpy 0.1735 0.1872 0.0843 0.0280 6.0164 5.9834 5.6258 0.6811 sitting-halfsphere 0.1005 0.0589 0.0389 0.0151 4.6490 2.8804 1.8836 0.6103 walking-static 0.3818 0.1339 0.0327 0.0293 6.3502 2.0833 0.8085 0.5500 walking-xyz 0.4360 0.2326 0.0651 0.1034 7.6669 4.3911 1.6442 2.3273 walking-rpy 0.4038 0.3584 0.2252 0.2143 7.0662 6.3898 5.6902 3.9555 walking-halfsphere 0.2628 0.1738 0.0527 0.1061 5.2179 4.2863 2.4048 2.2983
 Sequences RMSE of translational drift [m/s] RMSE of rotational drift [$^{\circ}$/s] DVO BaMVO SPW-VO Our Method DVO BaMVO SPW-VO Our Method sitting-static 0.0157 0.0248 0.0231 0.0112 0.6084 0.6977 0.7228 0.3356 sitting-xyz 0.0453 0.0482 0.0219 0.0132 1.4980 1.3885 0.8466 0.5753 sitting-rpy 0.1735 0.1872 0.0843 0.0280 6.0164 5.9834 5.6258 0.6811 sitting-halfsphere 0.1005 0.0589 0.0389 0.0151 4.6490 2.8804 1.8836 0.6103 walking-static 0.3818 0.1339 0.0327 0.0293 6.3502 2.0833 0.8085 0.5500 walking-xyz 0.4360 0.2326 0.0651 0.1034 7.6669 4.3911 1.6442 2.3273 walking-rpy 0.4038 0.3584 0.2252 0.2143 7.0662 6.3898 5.6902 3.9555 walking-halfsphere 0.2628 0.1738 0.0527 0.1061 5.2179 4.2863 2.4048 2.2983
RMSE of the ATE of camera pose estimation (m$^{-1}$)
 Sequences ORB-SLAM2 MR-SLAM SPW-SLAM SF-SLAM Our Method sitting-static 0.0082 – – 0.0081 0.0073 sitting-xyz 0.0094 0.0482 0.0397 0.0101 0.0090 sitting-rpy 0.0197 – – 0.0180 0.0162 sitting-halfsphere 0.0211 0.0470 0.0432 0.0239 0.0164 walking-static 0.1028 0.0656 0.0261 0.0120 0.0108 walking-xyz 0.4278 0.0932 0.0601 0.2251 0.0884 walking-rpy 0.7407 0.1333 0.1791 0.1961 0.3620 walking-halfsphere 0.4939 0.1252 0.0489 0.0423 0.0411
 Sequences ORB-SLAM2 MR-SLAM SPW-SLAM SF-SLAM Our Method sitting-static 0.0082 – – 0.0081 0.0073 sitting-xyz 0.0094 0.0482 0.0397 0.0101 0.0090 sitting-rpy 0.0197 – – 0.0180 0.0162 sitting-halfsphere 0.0211 0.0470 0.0432 0.0239 0.0164 walking-static 0.1028 0.0656 0.0261 0.0120 0.0108 walking-xyz 0.4278 0.0932 0.0601 0.2251 0.0884 walking-rpy 0.7407 0.1333 0.1791 0.1961 0.3620 walking-halfsphere 0.4939 0.1252 0.0489 0.0423 0.0411
 [1] Alain Chenciner, Jacques Féjoz. The flow of the equal-mass spatial 3-body problem in the neighborhood of the equilateral relative equilibrium. Discrete & Continuous Dynamical Systems - B, 2008, 10 (2&3, September) : 421-438. doi: 10.3934/dcdsb.2008.10.421 [2] Zvi Artstein. Invariance principle in the singular perturbations limit. Discrete & Continuous Dynamical Systems - B, 2019, 24 (8) : 3653-3666. doi: 10.3934/dcdsb.2018309 [3] Giuseppe Capobianco, Tom Winandy, Simon R. Eugster. The principle of virtual work and Hamilton's principle on Galilean manifolds. Journal of Geometric Mechanics, 2021, 13 (2) : 167-193. doi: 10.3934/jgm.2021002 [4] Navin Keswani. Homotopy invariance of relative eta-invariants and $C^*$-algebra $K$-theory. Electronic Research Announcements, 1998, 4: 18-26. [5] Yunsai Chen, Zhao Yang, Liang Ma, Peng Li, Yongjie Pang, Xin Zhao, Wenyi Yang. Efficient extraction algorithm for local fuzzy features of dynamic images. Discrete & Continuous Dynamical Systems - S, 2019, 12 (4&5) : 1311-1325. doi: 10.3934/dcdss.2019090 [6] Chun-Xiang Guo, Guo Qiang, Jin Mao-Zhu, Zhihan Lv. Dynamic systems based on preference graph and distance. Discrete & Continuous Dynamical Systems - S, 2015, 8 (6) : 1139-1154. doi: 10.3934/dcdss.2015.8.1139 [7] Dominique Zosso, Jing An, James Stevick, Nicholas Takaki, Morgan Weiss, Liane S. Slaughter, Huan H. Cao, Paul S. Weiss, Andrea L. Bertozzi. Image segmentation with dynamic artifacts detection and bias correction. Inverse Problems & Imaging, 2017, 11 (3) : 577-600. doi: 10.3934/ipi.2017027 [8] Amadeu Delshams, Josep J. Masdemont, Pablo Roldán. Computing the scattering map in the spatial Hill's problem. Discrete & Continuous Dynamical Systems - B, 2008, 10 (2&3, September) : 455-483. doi: 10.3934/dcdsb.2008.10.455 [9] Jian Zhai, Jianping Fang, Lanjun Li. Wave map with potential and hypersurface flow. Conference Publications, 2005, 2005 (Special) : 940-946. doi: 10.3934/proc.2005.2005.940 [10] Marian Gidea, Yitzchak Shmalo. Combinatorial approach to detection of fixed points, periodic orbits, and symbolic dynamics. Discrete & Continuous Dynamical Systems, 2018, 38 (12) : 6123-6148. doi: 10.3934/dcds.2018264 [11] Christine Bachoc, Gilles Zémor. Bounds for binary codes relative to pseudo-distances of $k$ points. Advances in Mathematics of Communications, 2010, 4 (4) : 547-565. doi: 10.3934/amc.2010.4.547 [12] Yunkyong Hyon, José A. Carrillo, Qiang Du, Chun Liu. A maximum entropy principle based closure method for macro-micro models of polymeric materials. Kinetic & Related Models, 2008, 1 (2) : 171-184. doi: 10.3934/krm.2008.1.171 [13] Iordanka N. Panayotova, Pai Song, John P. McHugh. Spatial stability of horizontally sheared flow. Conference Publications, 2013, 2013 (special) : 611-618. doi: 10.3934/proc.2013.2013.611 [14] Seyedeh Marzieh Ghavidel, Wolfgang M. Ruess. Flow invariance for nonautonomous nonlinear partial differential delay equations. Communications on Pure & Applied Analysis, 2012, 11 (6) : 2351-2369. doi: 10.3934/cpaa.2012.11.2351 [15] Xianchao Xiu, Ying Yang, Wanquan Liu, Lingchen Kong, Meijuan Shang. An improved total variation regularized RPCA for moving object detection with dynamic background. Journal of Industrial & Management Optimization, 2020, 16 (4) : 1685-1698. doi: 10.3934/jimo.2019024 [16] Amin Aalaei, Hamid Davoudpour. Two bounds for integrating the virtual dynamic cellular manufacturing problem into supply chain management. Journal of Industrial & Management Optimization, 2016, 12 (3) : 907-930. doi: 10.3934/jimo.2016.12.907 [17] Yaofeng Su. Almost surely invariance principle for non-stationary and random intermittent dynamical systems. Discrete & Continuous Dynamical Systems, 2019, 39 (11) : 6585-6597. doi: 10.3934/dcds.2019286 [18] Wacław Marzantowicz, Piotr Maciej Przygodzki. Finding periodic points of a map by use of a k-adic expansion. Discrete & Continuous Dynamical Systems, 1999, 5 (3) : 495-514. doi: 10.3934/dcds.1999.5.495 [19] Lok Ming Lui, Tsz Wai Wong, Wei Zeng, Xianfeng Gu, Paul M. Thompson, Tony F. Chan, Shing Tung Yau. Detection of shape deformities using Yamabe flow and Beltrami coefficients. Inverse Problems & Imaging, 2010, 4 (2) : 311-333. doi: 10.3934/ipi.2010.4.311 [20] Timothy Blass, Rafael De La Llave, Enrico Valdinoci. A comparison principle for a Sobolev gradient semi-flow. Communications on Pure & Applied Analysis, 2011, 10 (1) : 69-91. doi: 10.3934/cpaa.2011.10.69

2020 Impact Factor: 1.801

## Tools

Article outline

Figures and Tables