[feat] [fix] PCL Fitting fix and generalization
- Add PCL features module to use PCL normals. Compute PCL normals for cones and cylinder and enable the models.
- Add searchRadius parameter to compute normals for cone and cylinder models.
- Add radius limits for models using a radius.
- Add the property SamplesMaxDist in order to specify the maximum distance allowed when drawing random samples. An IsSamplesMaxDistUsed property is also added to enable SampleMaxDist. By default, this properly is disabled and SampleMaxDist is set to infinity.
- Add perpendicular and normal Plane models
- Add an option to project inliers on model (Not implemented for spheres)
- Add ouput options (Inliers or outliers can be used as output instead of all data)
- Add multi-instance fitting
- Add model constraints (a std::function can be given to the model to apply constraints.)
- Save model coefficients in a std::vector
- Add a vector to provide Indices to use only a part of the input pointcloud
- Generalize vtkPCLRansacModel to other sample consensus algorithms.
- Rename class for generalization
Edited by Julien Chaize