Advanced image analysis for particle image velocimetry. Part I: Processing individual PIV recordings

The article provides an overview of existing (digital) image evaluation methods for particle image velocimetry (PIV), which were developed in the past 20 years, and demonstrates their respective performance using actual PIV image recordings. The algorithms share the common goal of faithfully extracting the displacement field encoded by the motion of imaged particles from one image recording to the next using various implementations of mainly correlation-based pattern matching. Modern algorithms make use of multi-resolution coarse-to-fine processing strategies augmented with iterative, full image distortion in a predictor-corrector sense to increase both the spatial resolution and velocity dynamic range obtainable with PIV. The article reviews various processing schemes, required image interpolation methods as well as correlation signal enhancing image spatial or spectral filtering