Optimization Based Reconstruction for Digital In-Line Holographic Microscopy

High resolution three-dimensional imaging of micro-scopic samples can be obtained through non-invasive technique of Digital In-line Holographic Microscopy (DIHM). The paper offers a compressive sensing based reconstruction technique solved through Alternating Direction Method of Multipliers (ADMM) optimization. The reconstruction is supported with a physical constraint to improve the resolution of phase reconstruction. Morphological analysis of soft tissues is an important part of carcinoma detection. This work demonstrates the effectiveness of this type of reconstruction in phase retrieval from digital holograms of soft tissue. The results shows better resolution and improved Root Mean Square Error (RMSE) and Structural Similarity Index (SSIM).