张阳阳,黄英,郝超,郭小辉,刘平.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
基于织物拉伸传感器的手势映射系统
Gestures mapping system based on fabric strain sensor
  
DOI:
中文关键词:  拉伸传感器  织物  D H算法  反馈机制  手势映射系统
英文关键词:strain sensor  fabric  D H algorithm  feedback mechanism  gesture mapping system
基金项目:国家自然科学基金(61471155,61673369)项目资助
作者单位
张阳阳 (合肥工业大学大学电子科学与应用物理学院合肥230009 
黄英 (合肥工业大学大学电子科学与应用物理学院合肥230009 
郝超 (合肥工业大学大学电子科学与应用物理学院合肥230009 
郭小辉 (合肥工业大学大学电子科学与应用物理学院合肥230009 
刘平 (合肥工业大学大学电子科学与应用物理学院合肥230009 
AuthorInstitution
Zhang Yangyang School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China 
Huang Ying School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China 
Hao Chao School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China 
Guo Xiaohui School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China 
Liu Ping School of Electronic Science & Applied Physics, Hefei University of Technology, Hefei 230009, China 
摘要点击次数: 1231
全文下载次数: 1671
中文摘要:
      针对人机交互中主从手映射系统反馈信息缺乏、映射精确度及可穿戴性差等存在的缺点,提出一种基于织物可拉伸传感器以及具有反馈机制的主从手映射操控系统。以莱卡织物表面旋涂石墨烯/聚苯胺复合导电材料,有机硅导电银胶作为电极,研制一种全柔性织物可拉伸传感器。同时将该拉伸传感器布置于人手及机械手,构成2×5的拉伸传感阵列。结合主从手拉伸传感信息,通过改进D H算法,建立手势识别模型。利用BP神经网络对主从手拉伸传感器信息映射进行离线建模,结合在线优化算法,引入拉伸传感器信息反馈机制,实现主从手交互手势的高效、精准映射,建立具有反馈机制的手势映射系统。通过实验分析传感器的拉伸特性,并对有、无反馈机制的手势映射系统的映射精准度进行对比实验。结果表明,所设计的基于柔性织物拉伸传感器、具有反馈机制的主从手映射系统提升了手势映射的精准度,并兼具良好的穿戴性。
英文摘要:
      Current master slave hand mapping in the human computer interaction has shortcomings such as lack of feedback, poor mapping accuracy and poor wearability. A master slave mapping system based on fabric strain sensor and feedback mechanism is presented to address these issues. The strain sensitive unit is constituted of Lycra fabric surface spinning graphene/polyaniline composite conductive materials, in which the silver conductive adhesive is used as the electrode. The fabric strain sensor is layout in master slave hand to construct 2×5 array. The gesture recognition model is obtained by combining stretch sensor information and the improved D H algorithm. The BP neural network is used to model the information of the master slave strain sensor. Combining with the online optimization algorithm, the feedback mechanism of the strain sensor is introduced to realize the efficient and accurate mapping of the master slave hand gestures, and the gesture mapping with feedback mechanism is established. The strain characteristics of the sensors are tested and the mapping accuracy of the gesture mapping system with and without feedback mechanism is compared. The experimental results indicate that the master slave control system based on fabric strain sensor and feedback mechanism can improve mapping precision and wearability.
查看全文  查看/发表评论  下载PDF阅读器