The interrogation analysis of digital PlY images with low pixel resolution is investigated analytically. It is demonstrated
that the minimum required sampling rate for PlY images is about 1/4 of the sampling rate that matches
the optical bandwidth of PlY imaging optics; a further increase of the sampling rate does not improve the accuracy
of the estimated image covariance. The accuracy of a centroiding ( "center-of-mass" ) estimator and a curve-fitting
( "Gaussian-fit" ) estimator for the particle-image displacement is evaluated. It appears that the estimated displacement
is always biased, as a result of the fact that the image covariance is estimated from a finite number of samples.
The proposed correction procedure adequately compensates for this bias. The predicted performance of the displacement
estimators is in fair agreement with the results obtained from measurements of a linearly displaced test object.