Learning-based method for full phase reconstruction of biological samples in digital holographic microscopy

Authors

Raul Castaneda, University of MemphisFollow
Carlos Trujillo, Universidad EAFIT
Ana Doblas, University of Memphis

Abstract

A convolutional autoencoder for complete phase reconstruction in Digital Holographic Microscopy is reported. After proper training, this computationally efficient method reconstructs DHM holograms accurately when compared to the traditional approaches. This learning-based method is trained and validated with experimental samples of red blood cells.

Publication Title

2021 IEEE Photonics Conference, IPC 2021 - Proceedings

Recommended Citation

Castaneda, R., Trujillo, C., & Doblas, A. (2021). Learning-based method for full phase reconstruction of biological samples in digital holographic microscopy. 2021 IEEE Photonics Conference, IPC 2021 - Proceedings https://doi.org/10.1109/IPC48725.2021.9592907