<> <> This package implements point cloud registration, using the Normal Distributions Transform. Two basic classes are available - point to distributiuon registration (NDTMatcherP2D) and distribution to distribution registration (NDTMatcherD2D). The point to distribution algorithm is described [[http://aass.oru.se/Research/mro/publications/Magnusson_etal_2007-JFR-3D_Scan_Registration_for_Autonomous_Mining_Vehicles.html|here]] and the distribution to distribution [[http://aass.oru.se/Research/mro/publications/Stoyanov_etal_2012-IJRR_Fast_and_Accurate_Scan_Registration_through_Minimization_of_the_Distance_between_Compact_3D_NDT_Representations.html|here]]. == Documentation == Refer to CodeAPI for detailed description of the classes and interfaces. == Example == {{{ #!cplusplus #include #include int main() { double __res[] = {0.5, 1, 2, 4}; std::vector resolutions (__res, __res+sizeof(__res)/sizeof(double)); lslgeneric::NDTMatcherD2D matcherD2D (false,false,resolutions); Eigen::Transform T; pcl::PointCloud cloud1,cloud2; //... load information into cloud1 and cloud2 ... bool ret = matcherD2D.match2D(cloud1,cloud2,T); //we now have T, which minimizes |cloud1 - cloud2*T| lslgeneric::transformPointCloudInPlace(T,cloud2); } }}} == Nodes == ## AUTOGENERATED DON'T DELETE ## CategoryPackage