Full project title: “Comparative analysis and hybrid method proposal of Gerchberg-Saxton and Transport of Intensity Equation algorithms for phase reconstruction from defocused images in digital in-line holographic microscopy”. 

One of the main challenges of modern microscopy is the observation of transparent objects. Among the techniques enabling their imaging, phase imaging (PI) techniques provide high measurement accuracy, without the need to interfere with the object under study. PI methods rely on the estimation of the phase delay of light passing through the sample, usually using the interference of object and reference beams (interferometry, holography). Other PI methods are intensity based, including transport of intensity (TIE) and on-axis Gabor holography (GH). Both of them recover phase from defocused images and can be implemented in identical on-axis digital holographic microscopy (DHM) systems. The main difference between them is the defocusing distance, which for the TIE method should be smaller (around in-focus distance) than for GH (usually above 50 μm). Additionally, the number of images needed for correct phase reconstruction is an important factor. For TIE, at least 2 images with different defocus must be collected, while the GH method allows reconstruction from a single hologram. However, such reconstruction will be subject to the so-called twin-image effect, which may be minimized by collecting at least 2 axially separated holograms and then applying the iterative Gerchberg-Saxton (GS) algorithm. Despite the significant similarities between these methods, a comprehensive comparison of them has not yet been proposed. So far, several hybrid algorithm solutions combining the TIE and GS approaches have also been proposed. However, due to the limitation to one of the regimes (small defocus – TIE, or larger – GS), they have not allowed to unleash the full potential of hybrid operation of both methods. The goal of the INHOLO project is to conduct a comparison between the above-mentioned algorithms in order to determine their optimal working regimes and to propose the novel hybrid method, combining TIE with GS allowing for better reconstruction.

Principal Investigator: Prof. Michał Józwik

This section is currently being updated.