The idea is to have two separated full frame images with a short interframing time.
After the acquisition the two images are divided into sub-areas and cross- correlated, giving the direction and speed of the flow of interest for many points in space.
The first image is acquired and transferred into the interline registers in response to the trigger.
The second image is stored in the photosites and must wait for the first image to be read out before it can be transferred and read out (see Fig.). The typical light sources for PIV are Q-switched lasers or flashes with short light pulses, freezing most movements.
The short and well defined light pulses make it possible to place a light pulse on each side of the sensor gate pulse – the strobe blanking period is 500 ns.