Correlation-based PIV has the advantage over particle tracking algorithms in which each particle path is followed. In contrast, in correlation based PIV, the average motion of small group of particles contained in the interrogation spot is calculated by spatial auto correlation or cross correlation. Auto correlation is performed when images for both laser pulses are recorded on the same sensor, while in cross correlation, each pulse is collected into separate frames. Cross correlation calculation becomes faster in the frequency domain since the FFT algorithm is now applicable. There is directional ambiguity in auto correlation technique. Hence, in case of reversed flow, this technique is not suitable. The drawback can be eliminated using the cross correlation technique. Cross correlation allows us to use a small interrogation area compared to auto correlation and leads to a reduction of the random error due to spatial velocity gradients. An important condition involves depth of field of recording optics and laser light sheet thickness. Generally, depth of the field should not be smaller than the thickness of the light sheet in order to avoid imaging out-of-focus particle.

Figure 3.30: Cros Corellation analysis of a piv Image Pair

Figure 3.31: Computation of cross correlation using FFT.