A sensor for large strain deformation measurement with automated grid method based on machine vision

Abstract

The strain of the large deformation sample, such as rubber, can not be measured by the material testing machine because of the extensometer restriction for testing requirement. In order to solve this problem, a sensor for in-plane strain measurement of a large deformation sample based on machine vision was developed and verified for its application to rubber material characteristics testing. The system consists of a CMOS image sensor for image acquisition, a DaVinci DSP for image collection and processing, one S3C2440 chip for control function and improved automated grid method to achieve the desired efficiency and accuracy, which has been redesigned to be appropriate for embedded system. The calibration accuracy of this system is about 8 microns after decreasing distortion through camera calibration with the calibration board whose position precision is about 1.5 microns. The results of non-contact rubber tension test indicate that the system is reliable and stable,and the measurement range of strain exceeds 100%. This system has the potential in mechanical properties testing,industry measurement and other application areas. © 2013 Springer-Verlag Berlin Heidelberg.