基于人体动作识别的类人机器人动作模仿

doi:10.3969/j.issn.1000-1158.2021.09.03

摘要
图/表
参考文献(19)
相关文章 (15)

全文: PDF (2212 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要基于模仿学习以及人机交互技术,借助Kinect深度相机传感器对类人机器人上半身动作模仿问题进行了研究。首先,将改进的D-H模型应用于NAO机器人的手臂完成了手臂运动学模型的精确建立及求解,解决了传统建模方法两相邻关节平行时出现的奇异性问题。其次,提出了一种改进的基于深度图像的手势识别算法,完成了对于示教者手势的判定及模仿,与传统基于彩色图像的手势识别相比,不受光照影响的同时提升了识别准确率,改进算法的平均识别准确率达到96.2%。最后将NAO机器人作为试验平台的实验表明:NAO机器人对于示教者上半身动作的实时在线模仿运动轨迹平滑且稳定,并且在抓取实验中也显现出了较好的准确性。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	朱奇光
	董惠茹
	张孟颖

关键词 ：计量学, 类人机器人, 手势识别, 模仿学习, 人机交互, 改进D-H模型

Abstract：Based on imitation learning and human-computer interaction technology, Kinect depth camera sensor is used to study the upper body motion imitation of humanoid robots. Firstly, the modified D-H model is applied to the arms of the NAO robot to complete the accurate establishment and solution of the kinematics model of the arms, and solve the singularity problem when two adjacent joints are parallel. Secondly, an improved gesture recognition algorithm based on depth image is proposed, which judges and imitates the gestures of the teacher. Compared with the traditional gesture recognition based on color image, it is not affected by light and improves recognition accuracy of the system and the average recognition accuracy of the improved algorithm reaches 96.2%. Finally, the experiments using the NAO robot as a test platform show that the system enables the NAO robot to simulate the upper body movements of the teacher in real time, with smooth and stable motion trajectory, and it also shows good accuracy in the grasping experiment.

Key words： metrology imitation learning gesture recognition human-computer interaction humanoid robot modified D-H model

收稿日期: 2019-12-24 发布日期: 2021-09-24

PACS:	TB92
	TP242.6

基金资助:国家自然科学基金(61773333);河北省自然科学基金(F2016203245);河北省教育厅高等学校科技计划重点项目(ZD2018234)

作者简介: 朱奇光(1978-),男,浙江宁波,燕山大学副教授,主要从事机器视觉、智能移动机器人检测与控制方面的研究。Email:zhu7880@ysu.edu.cn

引用本文:

朱奇光,董惠茹,张孟颖. 基于人体动作识别的类人机器人动作模仿[J]. 计量学报, 2021, 42(9): 1136-1141.
ZHU Qi-guang,DONG Hui-ru,ZHANG Meng-ying. Humanoid Robot Motion Imitation Based on Human Posture Recognition. Acta Metrologica Sinica, 2021, 42(9): 1136-1141.

链接本文:

http://jlxb.china-csm.org:81/Jwk_jlxb/CN/10.3969/j.issn.1000-1158.2021.09.03 或 http://jlxb.china-csm.org:81/Jwk_jlxb/CN/Y2021/V42/I9/1136

［1］黄英, 陆伟, 赵小文, 等. 用于机器人关节的柔性压力触觉传感器设计［J］. 计量学报, 2012, 33 (6): 523-527.
Huang Y, Lu W, Zhao X W, et al. Design of flexble pressure tactile sensor for robot joint［J］. Acta Metro-logica Sinica, 2012, 33(6):523-527.
［2］Duy N T, Jan P. Model learning for robot control: a survey ［J］. Cognitive processing, 2011, 12 (4): 319-340.
［3］任瑜, 张丰, 郭志敏, 等. 一种通用的工业机器人位姿检测方法［J］. 计量学报, 2018, 39 (5): 615-621.
Ren Y, Zhang F, Guo Z M, et al. A general pose testing method of industrial robot ［J］. Acta Metrologica Sinica, 2018,39 (5): 615-621.
［4］Shahverdi P, Masouleh M T. A simple and fast geometric kinematic solution for imitation of human arms by a NAO humanoid robot ［C］//International Conference on Rob-otics and Mechatronics. Tehran, 2017: 572-577.
［5］Kondo Y, Takahashi Y. Real-time whole body imitation by humanoic robot based on particle filter and dimension reduction by autoencoder ［C］//2017 Joint 17th World Congress of International Fuzzy Systems Association and 9th International Conference on Soft Computing and Intelligent Systems. Otsu, Japan, 2017: 1-6.
［6］Alibeigi M, Rabiee S, Ahmadabadi M N. Inverse Kine-matics Based Human Mimicking System using Skeletal Tracking Technology ［J］. Journal of Intelligent and Robotic Systems, 2017, 85 (1): 27-45.
［7］Elbasiony R, Gomaa W. Humanoids skill learning based on real-time human motion imitation using Kinect ［J］. Intelligent Service Robotics, 2018, 11 (2): 149-169.
［8］Chen J X, Wang G W, Hu X, et al. Lower-body control of humanoid robot NAO via Kinect ［J］. Multimedia Tools and Applications, 2017, 77 (5): 10883-10898.
［9］于建均, 张远, 左国玉, 等. 基于质心校正补偿的仿人机器人模仿学习［J］. 北京工业大学学报, 2018, 44 (2): 193-199.
Yu J J, Zhang Y, Zuo G Y, et al. Humanoid Robot Imitation Learning Based on COM Correction and Compe-nsation ［J］. Journal of Beijing University of Technology, 2018, 44 (2): 193-199.
［10］Wen S H, Ma Z Y, Zhao Y S, et al. The study of NAO robot arm based on direct kinematics by using D-Hmethod［C］//2014 UKACC International Conference on Control. Loughborough, UK, 2014: 515-518.
［11］Li S Y, Caelli T, Ferraro M, et al. A novel bi-o-kinematic encoder for human exercise represent-ation and decomposition-Part 2: Robustness and optimisation ［C］//2013 International Conference on Control, Aut-omation and Information Sciences. Nha Trang, Vietnam, 2013: 30-35.
［12］张阳阳, 黄英, 郝超, 等. 基于织物拉伸传感器的手势映射系统［J］. 仪器仪表学报, 2017, 38 (10): 2422-2429.
Zhang Y Y, Huang Y, Hao C, et al. Gestures mapping system based on fabric strain sensor ［J］. Chinese Journal of Scientific Instrument, 2017, 38 (10): 2422-2429.
［13］Hang C C, Zhang R, Chen Z P, et al. Dynamic Gesture Recognition Method Based on Improved DTW Algorithm［C］//2017 International Conference on Industrial Infor-matics-Computing Technology, Intelegent Technology, Industrial Information Integration. Wuhan, China, 2017: 71-74.
［14］朱奇光, 张兴家, 陈卫东, 等. 基于颜色矩的改进尺度不变特征变换的移动机器人定位算法［J］. 计量学报, 2016, 37 (2): 118-122.
Zhu Q G, Zhang J X, Chen W D, et al. An Algorithm Research for Mobile Robot Localization Based on the Improved Scale Invariant Feature Transform of Color Moment ［J］. Acta Metrologica Sinica, 2016, 37 (2): 118-122.
［15］苗凯尧, 刘利芳, 张梦宇, 等. NAO机器人的语音交互与任务规划研究［J］. 电子测试, 2018, 391 (10): 61-62.
Miao K Y, Liu L F, Zhang M Y, et al. Research on Speech Interaction and Mission Planning of NAO Robot ［J］. Electronic Test, 2018, 391 (10): 61-62.
［16］赵春芳, 李江昊, 张大伟. 基于改进免疫遗传优化蚁群算法的移动机器人路径寻优研究［J］. 计量学报, 2019, 40 (3): 155-160.
Zhao C F, Li J H, Zhang D W. Robot Path Optimiz-ation Research Based on Improved Immune Genetic Optimization Ant Colony Algorithm ［J］. Acta Metrologica Sinica, 2019, 40 (3): 155-160.
［17］Akgun B, Cakmak M, Yoo J W, et al. Trajectories and keyframes for kinesthetic teaching: A human-robot interaction perspective ［C］// 2012 7th ACM/IEEE International Conference on Human-Robot Interaction. Boston, USA, 2012: 391-398.
［18］王松林, 徐文胜. 基于Kinect深度信息与骨骼信息的手指尖识别方法［J］. 计算机工程与应用, 2016, 52 (3): 169-173.
Wang S L, Xu W S. Fingertips recognition based on Kinect depth and skeleton data ［J］. Computer Engi-neering and Applications, 2016, 52 (3): 169-173.
［19］代沅兴, 于天彪, 杨建宇, 等. 一种新的基于轮廓特征的指尖及手指指向检测方法［J］. 东北大学学报 (自然科学版), 2018, 39 (7): 937-941.
Dai Y X, Yu T B, Yang J Y, et al. An Innovative Method for Fingertip and Finger Pointing Detection Based on Contour Features ［J］. Journal of Northeastern University, 2018, 39 (7):937-941.