AN IMAGE FILTER TECHNIQUE TO RELAX PARTICLE IMAGE VELOCIMETRY

The method particle image velocimetry (PIV) measures particle velocities in natural sciences and engineering based on the determination of the optical flow. Classical PIV requires capturing the particles twice with short delay and with very short exposure time which exceeds the frame rate and the shutter speed of typical imaging systems. Thus, the exposure time is realized with illumination bursts based on double pulse lasers and imaging systems capable of high frame rates are used for capturing the particles. In order to relax PIV measurements, a filter technique is presented that generates both snapshots out of only one single image. This image is captured with a standard imaging system with standard exposure times which relaxes the constrains of the Ply measurement setup significantly. The filter technique is based on spatial filters and its computational complexity is independent from the image content. It also enables the integration of the technique into embedded architectures for real-time PIV. The functionality is demonstrated by taking the example of a particle based manufacturing process and the measuring error caused by the filter technique is analyzed.