Using Python to do slam loop detection and graph optimization in window system

Because fast lio does not have a loop module, some existing github project projects that add loops have their own problems, so I saved the point cloud data and pose information of each node, and made a separate loop diagram optimization module by myself, which is completely decoupled from the LIO laser odometer module, and it should be simple enough to use a python code program file to achieve, and it can also be run in the window system. After some time of trying, I finally made this code tool.

Features:
Loosely coupled, it can be used in combination with any newly released odometer for pose optimization in the future

There are two modes, manually setting the point of the loop or automatically detecting the loop mode;

The code is simple and easy to run

Schematic diagram of automatic detection loop

The schematic diagram of optimizing the elevation error, red is the optimized track, effectively removes the elevation error:

The Open3D and NumPy modules are used.

If you have any questions or customized development requirements, you can contact me Email or WhatsApp : lonlonago@foxmail.com