I was trying to do the the colored point cloud registration demo in c++ (as opposed to the python example in the examples). I got it to compile and run but all I get back is the identity matrix
So I was looking at the code for the RegistrationColoredICP in ColoredICP.cpp. This function seems to prepare the target point cloud but then just pass it and the unchanged source pointcloud to the regular registration function.
auto target_c = InitializePointCloudForColoredICP(
target, geometry::KDTreeSearchParamHybrid(max_distance * 2.0, 30));
source, *target_c, max_distance, init,
Is this correct? I thought maybe instead I should prepare both target and source and then pass them to TransformationEstimationForColoredICP::ComputeTransformation?