Ling Hangjian, Katz Joseph
Appl Opt. 2014 Sep 20;53(27):G1-G11. doi: 10.1364/AO.53.0000G1.
This paper deals with two issues affecting the application of digital holographic microscopy (DHM) for measuring the spatial distribution of particles in a dense suspension, namely discriminating between real and virtual images and accurate detection of the particle center. Previous methods to separate real and virtual fields have involved applications of multiple phase-shifted holograms, combining reconstructed fields of multiple axially displaced holograms, and analysis of intensity distributions of weakly scattering objects. Here, we introduce a simple approach based on simultaneously recording two in-line holograms, whose planes are separated by a short distance from each other. This distance is chosen to be longer than the elongated trace of the particle. During reconstruction, the real images overlap, whereas the virtual images are displaced by twice the distance between hologram planes. Data analysis is based on correlating the spatial intensity distributions of the two reconstructed fields to measure displacement between traces. This method has been implemented for both synthetic particles and a dense suspension of 2 μm particles. The correlation analysis readily discriminates between real and virtual images of a sample containing more than 1300 particles. Consequently, we can now implement DHM for three-dimensional tracking of particles when the hologram plane is located inside the sample volume. Spatial correlations within the same reconstructed field are also used to improve the detection of the axial location of the particle center, extending previously introduced procedures to suspensions of microscopic particles. For each cross section within a particle trace, we sum the correlations among intensity distributions in all planes located symmetrically on both sides of the section. This cumulative correlation has a sharp peak at the particle center. Using both synthetic and recorded particle fields, we show that the uncertainty in localizing the axial location of the center is reduced to about one particle's diameter.
本文探讨了影响数字全息显微镜（DHM）在测量密集悬浮液中颗粒空间分布应用的两个问题，即实像与虚像的区分以及颗粒中心的精确检测。以往分离实场和虚场的方法包括应用多个相移全息图、组合多个轴向位移全息图的重建场以及分析弱散射物体的强度分布。在此，我们引入一种基于同时记录两个同轴全息图的简单方法，这两个全息图的平面彼此相隔一小段距离。该距离被选择为大于颗粒的拉长轨迹。在重建过程中，实像会重叠，而虚像会被全息图平面之间距离的两倍所位移。数据分析基于对两个重建场的空间强度分布进行关联，以测量轨迹之间的位移。该方法已应用于合成颗粒和2μm颗粒的密集悬浮液。相关性分析能够轻松区分包含1300多个颗粒的样品的实像和虚像。因此，当全息图平面位于样品体积内部时，我们现在可以实现对颗粒的三维跟踪。同一重建场内的空间相关性也用于改进颗粒中心轴向位置的检测，将先前引入的程序扩展到微观颗粒悬浮液。对于颗粒轨迹内的每个横截面，我们对该横截面两侧对称位置的所有平面内强度分布之间的相关性进行求和。这种累积相关性在颗粒中心处有一个尖锐的峰值。使用合成颗粒场和记录的颗粒场，我们表明将中心轴向位置定位的不确定性降低到约一个颗粒直径。
