An image-processing-based automated grid method is investigated to determine the method's displacement and strain accuracy limits, and how these limits are influenced by the choice of camera-calibration models. A CCD camera and a PC-based frame grabber are used to record grid spot motion, then ordering and centroiding are used to identity each spot and calculate their individual displacements. The displacements are fitted with a moving biquadratic surface, and the strains are obtained by analytical differentiation of that surface. Camera-calibration models which are considered include various combinations of image-perspective transformation, image stretching, and elliptical-lens distortion. The strain and displacement accuracy are explored through rigid-body motion and uniaxial tension tests. In the process, sensitivity to in-plane and out-of-plane rigid-body translation, and extreme sensitivity to in-plane rigid-body rotation (for non-synchronized frame grabbers) are confirmed. It is found that under the best conditions the displacement accuracy is .015 pixels and that the strain accuracy is 120 microstrain. Finally, the automated grid method is used to investigate the strains developed in an aluminum perforated strip subjected to uniaxial tension.
