Implement point cloud localization
Description
Point cloud localization is an important feature for a LiDAR-based stack.
This feature (i.e. NDT) should be implemented with a sufficient number of tests.
Posting implementation items from https://discourse.ros.org/t/autoware-hackathon-london/10270/9 :
-
Map publisher (pcd -> publish map): #136 (closed) -
Map (format) validation: #137 (closed) -
PointCloud2 -> NdtRepresentation: #138 (closed) -
Adapt voxel grid to make a plain NdtRepresentation: #138 (closed) -
Some kind of simple Point representation for scan #166 (closed) -
Newton’s method: #139 (closed) -
Simple line search: #139 (closed) -
Gradient and hessian computation in P2D optimization problem: #140 (closed) -
LocalizerBase: #141 (closed) -
initial guess stuff: #142 (closed) -
LocalizerBaseNode: #143 (closed) -
NdtLocalizer #141 (closed) -
NdtLocalizerNode #288 (closed)
Edited by Yunus Emre Çalışkan