Review of the state-of-the-art of in-line holographic microscopy

In-line Digital Holographic Microscopy (DHM) is an imaging technique that creates 3D images from holographic data. We categorize DHM reconstruction methods into direct, iterative, and machine learning based algorithms. Direct methods can be fast but often suffer from noise and low contrast due to the inherent twin-image aberration present in holographic reconstruction. Iterative methods improve accuracy and reduce noise but are generally more time consuming and computationally complex. Finally, machine learning methods can be extremely fast at inference but require large amounts of resources and training data. We provide a literary and illustrative summary of these methods (Figure 1), allowing for easy comparison.