This repository contains concise notes and code snippets summarizing the key ideas from the paper โOn-Manifold Preintegration for Real-Time Visual-Inertial Odometryโ by Forster et al.
- Markdown summaries of the main concepts:
- IMU measurement models
- Noise and bias handling
- On-manifold preintegration
- Factor graph formulation
- C++ code snippets illustrating:
- Preintegrated measurement computation
- Bias correction
- Residual and Jacobian construction
This repo provides a minimal, easy-to-read explanation of IMU preintegration targeted at:
- Students learning VIO / SLAM
- Engineers implementing VIO systems
- Anyone needing a lightweight reference without reading the full paper
This work is still in progress โ feel free to open issues or contribute improvements!