High-Speed 3D Imaging System Using Inline Phase Retrieval Holography: Application to Aerosol Particle Analysis in Laboratory

The application of inline holography for three-dimensional particle measurement has been accelerating due to technological advancements in imaging sensors and computational capabilities. In atmospheric sciences, this technique is employed to identify and characterize various types of aerosol particles within three-dimensional observation volumes. The next challenge in holography is to capture the dynamics of aerosol particles through high-resolution and high-speed imaging, thereby elucidating their interactions with the surrounding medium. We present an apparatus and software system for three-dimensional observation of aerosol particle trajectories, interactions with the surrounding medium, and particle collisions in laboratory experiments. This system focuses on water droplets ranging from tens to hundreds of micrometers in diameter. Our approach utilizes high-speed imaging with phase retrieval inline holography, coupled with ParticleHolography.jl, a GPU-accelerated Julia language package for 3D holographic reconstruction and particle analysis. The experimental setup employs single-fluid or dual-fluid spray nozzles with adjustable droplet size and flow rate, enabling the generation of homogeneous or heterogeneous water droplet sprays. This configuration aims to quantify collision efficiencies and other parameters within the aforementioned size range in future studies. Droplets are captured using two synchronized high-speed cameras in an inline setup, operating at 4000 frames per second. The resulting data undergoes phase retrieval and volume reconstruction processing using ParticleHolography.jl. ParticleHolography.jl, an open-source Julia package developed by the presenter, is designed for particle measurement using inline holography. It performs camera calibration, holographic reconstruction, particle detection, and trajectory tracking. Utilizing NVIDIA GPUs for computation, the package reduces processing time for the large amounts of data generated by high-speed hologram capturing. This presentation will provide specifications and performance evaluations of the apparatus, describe features and processing speeds of ParticleHolography.jl, and present initial experimental results. Additionally, we will discuss future research plans and potential applications of this system.