# Visual SLAM
###### tags: `archived` `research`

https://blog.csdn.net/MulinB/article/details/53421864

### Structure from Motion vs Visual SLAM[^Tombone]
* Structure from Motion (SfM):
* Offline
* no motion model
* A typical example: given a large collection of photos of a single outdoor structure, construct a 3D model of the structure and determine the camera's poses.

* Visual SLAM
* A real-time version of Structure from Motion (SfM).
* A "Geometric Method" in Computer Vision (rather than Learning-based Methods).
* Able to simultaneously build 3D maps of the world while tracking the location and orientation of the camera.
* SLAM has been slowly moving towards the low-power / real-time / single RGB camera mode of operation.
* [Example](https://youtu.be/GnuQzP3gty4)

### MonoSLAM
* First single camera-based SLAM research is done by Davison[^Davison]
* Previous works: restricted to smoothly moving robots and stereo vision.
* Real-time handling of uncertainty
* Motion model
* Davison in 2007: MonoSLAM: Real-Time Single Camera SLAM[^monoslam]
* [SceneLib2](https://github.com/hanmekim/SceneLib2)

### Other Resources
https://www.leiphone.com/news/201607/GLQj0wrjKD4eHvq5.html
https://www.cnblogs.com/gaoxiang12

[^Tombone]: http://www.computervisionblog.com/2016/01/why-slam-matters-future-of-real-time.html
[^Davison]:Davison. (2003). Real-time simultaneous localisation and mapping with a single camera. Proceedings Ninth IEEE International Conference on Computer Vision, 1403–1410 vol.2. https://doi.org/10.1109/ICCV.2003.1238654
[^monoslam]: Davison, A. J., Reid, I. D., Molton, N. D., & Stasse, O. (2007). MonoSLAM: Real-Time Single Camera SLAM. IEEE Transactions on Pattern Analysis and Machine Intelligence, 29(6), 1052–1067. https://doi.org/10.1109/TPAMI.2007.1049