即时定位与建图过程中位姿估计算法优化与设计研究
Research on Optimization and Design of Pose Estimation Algorithm in the Process for Simultaneous Localization and Mapping

DOI: 10.12677/CSA.2020.1012246, PP. 2331-2338

Keywords: 移动机器人，位姿估计，ICP算法，IDCPBoC算法，即时定位与建图
Mobile Robot, Pose Estimation, ICP Algorithm, IDCPBoC Algorithm, SLAM

Full-Text   Cite this paper   Add to My Lib

Abstract:

障碍物的检测和避免可以被视为设计移动机器人的中心问题，这项技术为机器人提供了可以在不熟悉的环境中穿越而不导致自身损坏。避障机器人可以检测路径上的障碍物并在不发生任何碰撞的情况下进行机动的设计。点集配准算法是移动机器人位姿估计和地图构建中的一个非常重要的算法，广泛应用于模型重建、多视角配准、即时定位与建图(SLAM)等领域，ICP算法已得到国内外研究学者的高度评价。本文比较了迭代最近点(ICP)和基于群集的双向迭代最近点(IDCPBoC)算法。基于以上两种算法，对机器人进行姿态估计，并对移动机器人避障实验过程进行了分析。密集传感器方法(DSM)可构建环境图，IDCPBoC算法可以更好地避免障碍，IDCPBoC算法的准确性和收敛性优于ICP算法，且IDCPBoC算法运行时间较少，因此实时性能也大大提高了。
Obstacle detection and avoidance can be considered as the central issue in designing mobile robots. This technology provides the robots with senses which can use to traverse in unfamiliar environments without damaging itself. In this paper, an Obstacle Avoiding Robot is designed which can detect obstacles in its path and maneuver around them without making any collision. It is a robot vehicle that works on Arduino Microcontroller and employs three ultrasonic distance sensors to detect obstacles. The point set registration algorithm is a very important algorithm in mobile robot pose estimation and map construction. It is widely used in the fields of model reconstruction, multi-view registration, simultaneous localization and mapping (SLAM), etc. The ICP algorithm has been studied at home and abroad. Scholars attach great importance to it. This paper compares the iterative closest point (ICP) and cluster-based iterative bidirectional closest point (IDCPBoC) algorithms. The two algorithms are used to estimate the pose of the robot; the obstacle avoidance experiment process of the mobile robot is analyzed; the environment map is constructed using the DSM (dense sensor method) method; the IDCPBoC algorithm can better avoid obstacles. The accuracy and convergence of the IDCPBoC algorithm are better than the ICP algorithm. Because the IDCPBoC algorithm runs less time, the real-time performance is also greatly improved.

References

[1]	张国良, 林志林, 姚二亮, 等. 考虑多位姿估计约束的双目视觉里程计[J]. 控制与决策, 2018, 33(6): 1008-1016.
[2]	张文安, 陈玉寅, 尹姝, 等. 一种基于多视觉传感器的刚体位姿估计方法[J]. 浙江工业大学学报, 2018, 46(6): 12-18.
[3]	李宇波, 朱效洲, 卢惠民, 等. 视觉里程计技术综述[J]. 计算机应用研究, 2012, 29(8): 2801-2805.
[4]	张毅, 陈起, 罗元. 室内环境下移动机器人三维视觉SLAM[J]. 智能系统学报, 2015(4): 615-619.
[5]	邱世聪, 罗意. 改进ICP算法的点云配准[J]. 河南科技, 2017(7): 40-42.
[6]	Mu？oz-Salinas, R., Marín-Jimenez, M.J. and Medina-Carnicer, R. (2018) SPM-SLAM: Simultaneous Localization and Mapping with Squared Planar Markers. Pattern Recognition, 86, 156-171. https://doi.org/10.1016/j.patcog.2018.09.003
[7]	李永锋, 张国良, 徐君. 基于Kinect的帧间配准改进ICP算法[J]. 电光与控制, 2016, 23(2): 56-60.
[8]	刘亚彬. 基于总体最小二乘的点云三维配准及改进的ICP算法研究[D]: [硕士学位论文]. 徐州: 中国矿业大学, 2016.
[9]	崔宝侠, 宋佳瑞. 未知环境下机器人避障及动态目标追踪[J]. 沈阳工业大学学报, 2018, 40(3): 55-61.
[10]	王艳, 孙向明, 熊珀艺, 等. 基于双目视觉的呼吸运动实时跟踪方法研究[J]. 中国生物医学工程学报, 2018, 37(1): 72-78.
[11]	刘昱岗, 王卓君, 刘艳芳, 等. 基于双目视觉图像的倒车障碍物检测预处理方法[J]. 重庆交通大学学报(自然科学版), 2018, 37(3): 92-98.
[12]	Mendes, M., Coimbra, A.P. and Crisostomo, M.M. (2018) Assis-Cicerone Robot with Visual Obstacle Avoidance Using a Stack of Odometric Data. IAENG International Journal of Computer Science, 45, 219-227.
[13]	Kalogeiton, V.S., Ioannidis, K., Sirakoulis, G.C., et al. (2019) Real-Time Active SLAM and Obstacle Avoidance for an Autonomous Robot Based on Stereo Vision. Cybernetics and Systems, 50, 239-260. https://doi.org/10.1080/01969722.2018.1541599
[14]	Tsai, C.L., Li, C.Y., Yang, G., et al. (2010) The Edge-Driven Dual-Bootstrap Iterative Closest Point Algorithm for Registration of Multimodal Fluorescein Angiogram Sequence. IEEE Transactions on Medical Imaging, 29, 636-649. https://doi.org/10.1109/TMI.2009.2030324
[15]	Kosuge, A., Yamamoto, K., Akamine, Y., et al. (2020) An SoC-FPGA-Based Iterative-Closest-Point Accelerator Enabling Faster Picking Robots. IEEE Transactions on Industrial Electronics. https://doi.org/10.1109/TIE.2020.2978722

Full-Text