RGB-D SLAM Algorithm in Indoor Dynamic Environments Based on Gridding Segmentation and Dual Map Coupling
AI Qinglin, WANG Wei, LIU Gangjiang
Key Laboratory of Special Purpose Equipment and Advanced Manufacturing Technology, Ministry of Education & Zhejiang Province, Zhejiang University of Technology, Hangzhou 310023, China
艾青林, 王威, 刘刚江. 室内动态环境下基于网格分割与双地图耦合的RGB-D SLAM算法[J]. 机器人, 2022, 44(4): 431-442.DOI: 10.13973/j.cnki.robot.210224.
AI Qinglin, WANG Wei, LIU Gangjiang. RGB-D SLAM Algorithm in Indoor Dynamic Environments Based on Gridding Segmentation and Dual Map Coupling. ROBOT, 2022, 44(4): 431-442. DOI: 10.13973/j.cnki.robot.210224.
Abstract:To deal with low pose estimation accuracy and poor mapping performance of existing RGB-D SLAM (simultaneous localization and mapping) systems in indoor dynamic environments, an RGB-D SLAM algorithm with gridding segmentation and dual map coupling is proposed. Based on homography motion compensation and bidirectionally compensated optical flow method, gridding motion segmentation is achieved according to the geometrical connectivity and clustering result on depth images. Meanwhile, the speed of the algorithm in the segmentation process is ensured. Only features in the static region are used to estimate the camera position by minimizing the reprojection error. By combining camera poses, RGB-D images, and gridding motion segmentation images, the sparse point cloud map and the static octree map of the scene are constructed simultaneously and coupled. The static map points are filtered out with the gridding segmentation method and the octree map based ray traversal method on the keyframes, and thus the sparse point cloud map is updated to ensure the positioning accuracy. Experimental results on public datasets and in actual dynamic scenes show that the proposed method can effectively improve the accuracy of camera pose estimation in indoor dynamic scenes and achieve the real-time construction and update for static octree maps of scenes. Moreover, the proposed method can run in real-time on common CPU hardware platforms without additional computational resources, such as GPU.
[1] Rosen D M, Doherty K J, Espinoza A T, et al.Advances in inference and representation for simultaneous localization and mapping[J].Annual Review of Control, Robotics, and Autonomous Systems, 2021, 4:215-242. [2] Saputra M R U, Markham A, Trigoni N.Visual SLAM and structure from motion in dynamic environments:A survey[J].ACM Computing Surveys, 2018, 51(2):1-36. [3] Sun Y X, Liu M, Meng M Q H.Motion removal for reliable RGB-D SLAM in dynamic environments[J].Robotics and Autonomous Systems, 2018, 108:115-128. [4] Wang Y B, Huang S D.Towards dense moving object segmentation based robust dense RGB-D SLAM in dynamic scenarios[C]//13th International Conference on Control, Automation, Robotics&Vision.Piscataway, USA:IEEE, 2014.DOI:10.1109/ICARCV.2014.7064596. [5] Lee S, Son C Y, Kim H J.Robust real-time RGB-D visual odometry in dynamic environments via rigid motion model[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems.Piscataway, USA:IEEE, 2019.DOI:10.1109/IROS40897.2019.8968208. [6] Sun Y X, Liu M, Meng M Q H.Improving RGB-D SLAM in dynamic environments:A motion removal approach[J].Robotics and Autonomous Systems, 2017, 89:110-122. [7] 高成强,张云洲,王晓哲,等.面向室内动态环境的半直接法RGB-D SLAM算法[J].机器人, 2019, 41(3):372-383.Gao C Q, Zhang Y Z, Wang X Z, et al.Semi-direct RGB-D SLAM algorithm for dynamic indoor environments[J].Robot, 2019, 41(3):372-383. [8] 魏彤,李绪.动态环境下基于动态区域剔除的双目视觉SLAM算法[J].机器人, 2020, 42(3):336-345.Wei T, Li X.Binocular vision SLAM algorithm based on dynamic region elimination in dynamic environment[J].Robot, 2020, 42(3):336-345. [9] 张慧娟,方灶军,杨桂林.动态环境下基于线特征的RGBD视觉里程计[J].机器人, 2019, 41(1):75-82.Zhang H J, Fang Z J, Yang G L.RGB-D visual odometry in dynamic environments using line features[J].Robot, 2019, 41(1):75-82. [10] 艾青林,刘刚江,徐巧宁.动态环境下基于改进几何与运动约束的机器人RGB-D SLAM算法[J].机器人, 2020, 43(2):167-176.Ai Q L, Liu G J, Xu Q N.An RGB-D SLAM algorithm for robot based on the improved geometric and motion constraints in dynamic environment[J].Robot, 2020, 43(2):167-176. [11] Bescos B, Facil J M, Civera J, et al.DynaSLAM:Tracking, mapping, and inpainting in dynamic scenes[J].IEEE Robotics and Automation Letters, 2018, 3(4):4076-4083. [12] He K M, Gkioxari G, Dollár P, et al.Mask R-CNN[C]//IEEE International Conference on Computer Vision.Piscataway, USA:IEEE, 2017.DOI:10.1109/ICCV.2017.322. [13] Yu C, Liu Z X, Liu X J, et al.DS-SLAM:A semantic visual SLAM towards dynamic environments[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems.Piscataway, USA:IEEE, 2018.DOI:10.1109/IROS.2018.8593691. [14] Badrinarayanan V, Kendall A, Cipolla R.SegNet:A deep convolutional encoder-decoder architecture for image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12):2481-2495. [15] Zhong F W, Wang S, Zhang Z Q, et al.Detect-SLAM:Making object detection and SLAM mutually beneficial[C]//IEEE Winter Conference on Applications of Computer Vision.Piscataway, USA:IEEE, 2018.DOI:10.1109/WACV.2018.00115. [16] Rother C, Kolmogorov V, Blake A."GrabCut":Interactive foreground extraction using iterated graph cuts[J].ACM Transactions on Graphics, 2004, 23(3):309-314. [17] 姚二亮,张合新,宋海涛,等.基于语义信息和边缘一致性的鲁棒SLAM算法[J].机器人, 2019, 41(6):751-760.Yao E L, Zhang H X, Song H T, et al.Robust SLAM algorithm based on semantic information and edge consistency[J].Robot, 2019, 41(6):751-760. [18] Redmon J, Farhadi A.YOLOv3:An incremental improvement[DB/OL].(2018-04-08)[2020-03-21].https://arxiv.org/abs/1804.02767v1. [19] Liu H J, Liu G L, Tian G H, et al.Visual SLAM based on dynamic object removal[C]//IEEE International Conference on Robotics and Biomimetics.Piscataway, USA:IEEE, 2019.DOI:10.1109/ROBIO49542.2019.8961397. [20] Campos C, Elvira R, Rodríguez J J G, et al.ORB-SLAM3:An accurate open-source library for visual, visual-inertial and multi-map SLAM[J].IEEE Transactions on Robotics, 2021, 37(6):1874-1890. [21] Rublee E, Rabaud V, Konolige K, et al.ORB:An efficient alternative to SIFT or SURF[C]//International Conference on Computer Vision.Piscataway, USA:IEEE, 2011.DOI:10.1109/ICCV.2011.6126544. [22] Arthur D, Vassilvitskii S.K-means++:The advantages of careful seeding[C]//8th Annual ACM-SIAM Symposium on Discrete Algorithms.New York, USA:ACM, 2007:1027-1035. [23] Khoshelham K, Elberink S O.Accuracy and resolution of Kinect depth data for indoor mapping applications[J].Sensors, 2012, 12(2):1437-1454. [24] Zennaro S, Munaro M, Milani S, et al.Performance evaluation of the 1st and 2nd generation Kinect for multimedia applications[C]//IEEE International Conference on Multimedia and Expo.Piscataway, USA:IEEE, 2015.DOI:10.1109/ICME.2015.7177380.