Design and development of a new strain measuring method based on smartphone and machine vision

Botao Xie,Xixian Chen,Mingwei Ding,Guangyi Zhou,Xuefeng Zhao

DOI: https://doi.org/10.1016/j.measurement.2021.109724

IF: 5.6

2021-01-01

Measurement

Abstract:In this study, a new strain measuring method was designed based on smartphone and machine vision algorithms, while the automatic calibration of the pixel size of images taken by the smartphone was achieved with the help of circle detection. A piston-type sensor to complement this method was designed, and the static and dynamic experimental results of the sensor matched well with those from FBG. Meanwhile, the sensor was tested with temperature compensation to obtain an empirical equation for the effect of temperature on the sensor. When compared with the FBG data, the static data of the sensor had the mean error of less than 4 mu epsilon and the standard error of the mean (SEM) of 1.9 mu epsilon, while the dynamic data had the mean error is less than 4.3 mu epsilon and the SEM is 5.4 mu epsilon. The results confirm that the sensor has great potential in structural health monitoring.