Autofocusing Method for Multifocal Holograms Based on Connected Domain Analysis

Kaiqi Lang,Jiaqing Qiang,Yuyi Qiu,Xiaoping Wang

DOI: https://doi.org/10.1016/j.optlaseng.2024.108624

IF: 5.666

2025-01-01

Optics and Lasers in Engineering

Abstract:Holography is widely used for imaging plankton because of its large depth of field. Autofocusing is a critical step in holographic imaging, where the optimal focal plane is determined by locating the extremum of the evaluation function curve. However, most of these methods perform auto-focusing on single focus holograms. To apply these methods to multi-focus holograms, the hologram must be subjected to object segmentation. It is difficult to achieve good results when focusing on holograms with small objects and large depth of field. Therefore, we propose a multi-focus autofocus method based on connected domain analysis (CDA), which allows focusing on multiple objects at different focal lengths simultaneously without image segmentation. Each object generates a focus evaluation curve, resulting in high focusing accuracy. We set up an in-line holographic optical system with a 30 mm depth of field and recorded holograms of Nauplius, Cladocera, Rotifera, Copepods, Noctiluca scintillans and real seawater samples. We demonstrated the robustness of the CDA algorithm for different target types and its effectiveness in focusing multi-focus images by applying it to holograms of six types of samples.