首页|基于残差网络的光触觉传感器触觉信息提取方法研究

基于残差网络的光触觉传感器触觉信息提取方法研究

扫码查看
光触觉传感器是一种通过摄像头或光纤采集光学图像,并通过图像处理方法提取触觉信息的传感器.为解决基于柔性光栅的光触觉传感器从图像到触觉信息的映射问题,设计了一种基于残差网络的触觉信息提取方法,通过探索不同的数据采集方式和图像预处理方法,对触点坐标和法向力的预测精度分别达到了 60 μm和 3 mN,并讨论了网络预测结果受不同因素的影响.为基于结构色的光触觉传感器提供了一种高效、高精度的触觉提取模式.
Research on Haptic Information Extraction Methods for Optical Tactile Sensors Based on Residual Network
Optical tactile sensor is a kind of sensor that collects optical images by camera or optical fiber and extracts haptic information by image processing method.To solve the problem of mapping from image to haptic information for a flexible grating-based optical tactile sensor,a set of haptic information extraction method based on residual network was designed,By exploring different data acquisition methods and image preprocessing methods,the prediction accuracy of the contact coordinates and normal force reaches 60 μm and 3 mN,and the effect of different factors on network prediction was discussed.This provides an efficient and high-precision haptic extraction mode for structured color based optical tactile sensors.

flexible gratingoptical tactile sensorresidual networkhaptic information extraction

闫春飞、邱宇泽、张焱、董二宝

展开 >

中国科学技术大学工程科学学院

柔性光栅 光触觉传感器 残差网络 触觉信息提取

2024

仪表技术与传感器
沈阳仪表科学研究院

仪表技术与传感器

CSTPCD北大核心
影响因子:0.585
ISSN:1002-1841
年,卷(期):2024.(11)