Digital image processing and pattern recognition have become
important tools in experimental stress analysis. Fringe interpretation and
digital autocorrelation are the two prominent examples where digital image
processing has significantly influenced experimental research. This
paper discusses the use of digital image processing in concert with the
grid method. Presented here is a start-to-finish analysis of grid point tracking
with image acquisition hardware and pattern recognition software. A
mathematical model, which is based on experimentally controllable parameters,
is used to simulate the video camera and analog-to-digital conversion.
A simple centroid algorithm is used to track the grid points. The
model results, which are reaffirmed by rigid body motion tests, show that
the fill factor, signal-to-noise ratio, and grid spot diameter govern the
accuracy of this automated grid method. Finally, the model shows that
the intensity profile of the individual grid spots has a significant impact
on the maximum allowable grid frequency, camera field of view, and the
overall aceuracy of automated grid methods.