基于双光场相机的高分辨率光场三维PIV技术

High resolution volumetric light field particle image velocimetry with dual plenoptic cameras

光场相机粒子图像测速(Light Field Particle Image Velocimetry,LF-PIV)是一种近几年新发展起来的流动测试手段,能够仅通过单个光场相机测量3D-3C瞬态速度场,简化了三维流场测量的实验复杂度,特别是能实现受限空间的三维速度场测量。然而这一技术尚存在一些不足:由于光场相机沿景深方向的空间分辨率较低,沿该方向的速度测量精度低于垂直于景深方向的测量精度。本文尝试从硬件角度入手,发展一种双光场相机流动测试技术,通过增大对示踪粒子的观察视角,来提高光场三维测量系统沿景深方向的空间分辨率。基于乘积代数迭代技术(Multiplicative Algebraic Reconstruction Technique,MART),开发了针对双光场相机的粒子三维重构算法。分别利用直接数值模拟(Direct Numerical Simulation,DNS)水射流的数字合成图像与低速水射流涡环的实验图像,将双光场相机的测量结果与单光场相机的测量结果进行对比分析研究。结果表明双光场相机与单光场相机相比显著提高了相机沿景深方向的测量精度。

Light field particle image velocimetry (LF-PIV) was recently introduced to measure the three-dimension three-component (3D-3C) velocity field with just single camera. The main drawback of this configuration is the decreased measurement accuracy in the out-of-plane dimension due to limited spatial resolution. This work presents a solution with dual-camera arrangement to broaden the viewing angle. Based on multiplicative algebraic reconstruction technique (MART), a novel particle reconstruction algorithm is developed for dual or multiple light field camera PIV system. In addition, the synthetic images of DNS jet data and experimental images of a vortex-ring are utilized to draw a comparison between the velocity measurement results of a single light field camera and dual light field cameras. It is determined that the accuracy of dual-camera configuration in the depth direction has been significantly improved when compared to single light-field camera system.