Open3D (C++ API)
0.18.0+252c867
|
#include <RobustKernel.h>
Public Member Functions | |
RobustKernel (const RobustKernelMethod type=RobustKernelMethod::L2Loss, const double scaling_parameter=1.0, const double shape_parameter=1.0) | |
Data Fields | |
RobustKernelMethod | type_ = RobustKernelMethod::L2Loss |
Loss type. More... | |
double | scaling_parameter_ = 1.0 |
Scaling parameter. More... | |
double | shape_parameter_ = 1.0 |
Shape parameter. More... | |
Base class that models a robust kernel for outlier rejection. The virtual function Weight(double residual); must be implemented in derived classes. This method will be only difference between different types of kernels and can be easily extended.
The kernels implemented so far and the notation has been inspired by the publication: "Analysis of Robust Functions for Registration Algorithms", Philippe Babin et al.
We obtain the correspondendent weights for each residual and turn the non-linear least-square problem into a IRSL (Iteratively Reweighted Least-Squares) problem. Changing the weight of each residual is equivalent to changing the robust kernel used for outlier rejection.
The different loss functions will only impact in the weight for each residual during the optimization step. For more information please see also: “Adaptive Robust Kernels for Non-Linear Least Squares Problems”, N. Chebrolu et al. The weight w(r) for a given residual r
and a given loss function p(r)
is computed as follow: w(r) = (1 / r) * (dp(r) / dr) , for all r Therefore, the only impact of the choice on the kernel is through its first order derivate.
GeneralizedLoss Method is an implementation of the following paper: @article{BarronCVPR2019, Author = {Jonathan T. Barron}, Title = {A General and Adaptive Robust Loss Function}, Journal = {CVPR}, Year = {2019} }
|
inlineexplicit |
double open3d::t::pipelines::registration::RobustKernel::scaling_parameter_ = 1.0 |
Scaling parameter.
double open3d::t::pipelines::registration::RobustKernel::shape_parameter_ = 1.0 |
Shape parameter.
RobustKernelMethod open3d::t::pipelines::registration::RobustKernel::type_ = RobustKernelMethod::L2Loss |
Loss type.