[1]黄海明,吴林源,林俊豪,等.基于内嵌型光学弯曲传感器的软体手感知[J].深圳大学学报理工版,2019,(No.3(221-346)):237-244.[doi:10.3724/SP.J.1249.2019.03229]
 HUANG Haiming,WU Linyuan,LIN Junhao,et al.The perception of soft hand based on embedded type optical curvature sensor[J].Journal of Shenzhen University Science and Engineering,2019,(No.3(221-346)):237-244.[doi:10.3724/SP.J.1249.2019.03229]
点击复制

基于内嵌型光学弯曲传感器的软体手感知()
分享到:

《深圳大学学报理工版》[ISSN:1000-2618/CN:44-1401/N]

卷:
期数:
2019年No.3(221-346)
页码:
237-244
栏目:
【电子与信息科学】
出版日期:
2019-05-20

文章信息/Info

Title:
The perception of soft hand based on embedded type optical curvature sensor
作者:
黄海明1吴林源1林俊豪1方斌2孙富春12
1)深圳大学信息工程学院,广东深圳 518060;2)清华大学智能技术与系统国家重点实验室,北京 100084
Author(s):
HUANG Haiming1 WU Linyuan1 LIN Junhao1 FANG Bin1 and SUN Fuchun12
1) College of Information Engineering, Shenzhen University, Shenzhen 518060, Guangdong Province, P.R.China;2) Department of Computer Science and Technology, Tsinghua University, Beijing 100084, P.R.China
关键词:
智能机器人柔性传感软体机器人光导纤维弯曲感知姿态识别
Keywords:
Intelligent robot flexible sensor soft robotic hand optical fiber curvature perception posture recognition
分类号:
TP242.6
DOI:
10.3724/SP.J.1249.2019.03229
文献标志码:
A
摘要:
软体仿人手的感知能力是实现友好交互和灵巧操作的重要条件.柔性的光学弯曲传感器由LED、光导纤维和光敏传感器组成,易嵌于软体手指内.经粗糙化的光导纤维可产生光线溢出,通过测量光能损耗量可以获得软体手指的弯曲度;利用归一化标定软体手指初始(伸直)和最大弯曲状态的采样电压值,对归一化后的电压值进行均匀、非均匀和滞环分段,并融合五指弯曲信息实现连续运动检测、动作姿态和物体大小识别.实验结果表明,光学弯曲传感器可检测连续正弦运动状态;正确识别多种手势和抓握姿态;也可识别物体大小,结合滞环分段方法,解决单点分类切换存在振荡不足的问题,提高识别率.这种光学式柔性弯曲传感器具有电磁免疫、易于内嵌的优势,在软体手和软体执行器中具有重要的应用价值.
Abstract:
It is important that the perception of soft humanoid hand is essential to come true friendly interactive and dexterous operation. Flexible optical curvature sensor is comprised of LED, optical fiber and photo-sensitive sensor, and easy to be embedded into the soft finger. The roughened optical fiber is able to produce light overflow. The bend degree of soft finger is obtained by detecting the loss of light energy. The abilities of continuous movement detection, action gestures and object size recognitions are implemented by the steps: calibrate the initial (unbend) and maximum bending statuses of soft fingers with normalization method; segment handing the post-normalization data by uniform, nonuniform and hysteresis method; merge together the information of five fingers. The experiment results indicate that the optical type curvature sensor is able to detect the continuous sinusoidal movement status, recognize correctly multiple types of gesture and grasping posture, recognize object size, and the recognition rate is improved and the oscillation is reduced owe to the hysteresis segment method. This optical type curvature sensor has the advantages of insensitive to electromagnetic and easy to be embedded. It is important in the application for soft hand or soft actuation.
更新日期/Last Update: 2019-04-22