Traditional: velodyne quanergy
Traditional lowcost: 思岚科技 雷神智能
Solid state: leddartech innoviz
https://data.kitware.com/#collection/5b7f46f98d777f06857cb206/folder/5b7f47568d777f06857cb208
Lord: https://www.microstrain.com/inertial/3DM-CV5-10
Xsens
cartographer-project/cartographer_ros#725
https://github.com/ytgcljj/ndt_localizer
VLP16: cartographer-project/cartographer_ros#259
KITTI: cartographer-project/cartographer_ros#264 (comment)
cartographer-project/cartographer#690
cartographer-project/cartographer_ros#628
cartographer-project/cartographer_ros#354
cartographer-project/cartographer_ros#332
cartographer-project/cartographer_ros#706
cartographer-project/cartographer#255
https://homepages.inf.ed.ac.uk/rbf/HIPR2/thin.htm
extropolator: pose only / odometry / correlative scan match
scan match: hit only / hit+miss matching?
Different purpose: pure localization / online SLAM / offline mapping
Scan frequency: 20HZ?
Rviz config: view submaps when running
Odometry: bicycle or rear differential?
Pure localization
Image registration
explaination: https://zhuanlan.zhihu.com/c_131391131
Pros: Accurate, even without IMU/LoopClosure
Cons: 1) Error occurs sometimes during tests. 2) How to remove dynamic objects?
video: https://www.youtube.com/watch?v=O3tz_ftHV48
Pros: Fast enough for Jetson TX2
Cons: Need IMU? Small errors during tests. LoopClosure seems not working well?
GVD: may use boost and merge linesegments?
Solver: use Ceres instead of own code?
http://www.inf.u-szeged.hu/~palagyi/skel/skel.html
https://link.springer.com/article/10.1007/s12647-018-0289-x
http://www.docin.com/p-521936408.html