Documentation for Differentiable Numeric Optical Imaging Simulation (DNOIS)

Introduction

Differentiable Numeric Optical Imaging Simulation (DNOIS) is a Python library developed for simulating and optimizing diverse imaging systems, which involves both optical components and optoelectronic sensors. The aims of DNOIS include:

  • Provides precise, standard, efficient and usable implementation for imaging simulation.

  • Covers whole imaging pipeline in computational photography/imaging, from representation of scenes to output signal of simulated sensor.

  • Simulates imaging in a differentiable manner, which enables end-to-end optimization.

  • Supports diverse application scenarios and flexible options of optical systems and algorithms.

Documentation

This documentation consists of four parts:

  • Getting Started: A quick start guide for users to get started with DNOIS.

  • Tutorials: A series of demos to help users understand the usage of DNOIS in a practical way.

  • User Guide: An in-depth guide explaining key components in DNOIS, detailed but slightly tedious.

  • API Reference: A comprehensive reference for the API of DNOIS.

Indices and tables