Computational Optical Imaging

This repository contains a collection of Jupyter notebooks developed as part of the Computational Optical Imaging course at EPFL. In this course, we learnt how to use computational tools to simulate optical systems and combine them with neural networks that process optical images.

---

Topics covered:

1. Optical wave propagation

   • Free-space propagation
   • Beam propagation method (BPM)
   • Thin transparencies: lenses and gratings
   • Imaging systems
   • Digital holography
   • Computer-generated holograms (CGH)

2. Multi-layer networks

   • Neural architectures for image enhancement and inversion
   • Training principles for optical simulations

3. Microscopy

   • UNet-based super-resolution
   • Digital staining
   • Phase retrieval from intensity images

4. Scattering media

   • Phase conjugation
   • Matrix methods
   • DNNs for focusing and imaging through multimode fibers (MMFs)
   • Ptychography

5. Inverse scattering

   • Optical diffraction tomography
   • MaxwellNet and inverse-scattering reconstruction

---

Exercises

Notebook	Title / Topic
GX_01.ipynb	Free space propagation
GX_02.ipynb	Angular Spectrum Method, diffraction
GX_03.ipynb	GRIN, BPM
GX_04.ipynb	Zernike phase shift contrast microscope
GX_05.ipynb	Coherent & Incoherent, Fourier optics
GX_06.ipynb	NN for optics
GX_07.ipynb	Feature size
GX_08.ipynb	Gerchberg Saxton algorithm
GX_09.ipynb	Waveguides
GX_10.ipynb	Holography
GX_11.ipynb	Fibers

---

Example from one of the notebook

Simulated propagating modes in a fiber.