Rapid cm-scale 3D PIV/PTV with plenoptic macro photography

Multi-camera-based flow description in Eulerian and Lagrangian frames of reference at mm and cm scales is usually challenging due to shallow depth of field (DOF) from macro lenses and calibration procedures. We will describe a rapid 3D particle tracking system developed for cm-scale volume applications. With the use of a microlens array (MLA) in the intermediate plane between the macro lens and camera sensor, light field information can be obtained by a single camera cage system without stereo calibration. A fast ray intersection and cloud point classification method with GPU acceleration is applied to reconstruct 3D particle position in a uniformly discretized image space and then mapped to non-uniform objective space based on macro lens properties. With atomic operation within the onboard memory of GPU, the processing time per image can be under one second. Lagrangian particle tracking Crocker-Grier method is used for particle linking to characterize time-resolved flow field.