Digital image correlation : from integer pixel to sub-pixel

Abstract

In collaboration with the advancement of computer technology and digital image acquisition (digital imaging), Digital Image Correlation (DIC) technique is getting more common in analyzing strain/deformation on the surface.

This report investigates the basic method of digital image correlation (DIC), limitation, and implementation on strain analysis. Simple DIC could predict up to 1 pixel accuracy. To get more accuracy, there have been several famous techniques based on previous research such as Gradient-Based method, Newton-Raphson Method, and Curve-Fitting method. This report gives comparison on performance of those techniques as proof of concept and its importance to strain/deformation analysis.