Li Shengfu, Zhao Yu, Ye Yan
J Opt Soc Am A Opt Image Sci Vis. 2019 Aug 1;36(8):1395-1401. doi: 10.1364/JOSAA.36.001395.
In-line particle holography suffers from speckle noise, and great effort has been made to alleviate it by designing post-processing algorithms. Recently, we proposed a novel approach which mitigates it by increasing the signal-to-noise ratio (SNR). The approach enhances the SNR by combining several holograms captured under different illumination angles. Accurate knowledge of the illumination angles is essential for a high-quality reconstruction, and thus requires system calibration which is often time- and labor-intensive. Here, to eliminate the need for intensive pre-calibration, we propose a self-calibrated approach that estimates the illumination angles directly from experimentally collected holograms without requiring any additional hardware or captured images. We demonstrate our method for correcting the illumination misalignment with both numerical simulation and experimental data.
在线粒子全息术存在散斑噪声问题，人们已付出巨大努力通过设计后处理算法来减轻这种噪声。最近，我们提出了一种新颖的方法，该方法通过提高信噪比（SNR）来减轻散斑噪声。该方法通过组合在不同照明角度下捕获的多个全息图来提高信噪比。准确了解照明角度对于高质量重建至关重要，因此需要进行系统校准，而这通常既耗时又费力。在此，为了消除密集预校准的需求，我们提出了一种自校准方法，该方法可直接从实验收集的全息图中估计照明角度，而无需任何额外的硬件或捕获图像。我们通过数值模拟和实验数据展示了我们用于校正照明失准的方法。
