J054014 デコンボリューションを用いたディジタルホログラフィ再生像の深さ方向伸び抑制(〔J054-01〕流れの高次元化可視化に向けて(1))

Abstract

This paper presents a deconvolution technique for suppressing the image elongation in depth in reconstruction of digital holography. Target objects are small particles distributed in 3D space and the present technique is applied to the measurement of particle positions. The 3D spatial distribution of reconstructed light wave for a single particle is used as Point Spread Function (PSF) in deconvolution process. Though reconstructed particle images are elongated along depth direction in digital holography, the elongation is much suppressed according to the PSF to easily and accurately detect the positions of particles recorded in hologram patterns. Numerical simulations are carried out for evaluating the performance of the present technique for two models of a single particle and multi-particles. From the numerical simulation results for single-particle model, it is shown that an elongated particle image is the most successfully converted to a compact one for complex amplitude distribution that is computed from a hologram recorded with a phase-shifting method. The simulation for multi-particle model is performed up to 100 particles and the RMS error in detected particle position is found to be less than that by conventional method without convolution.