Hello,
After generating the mesh I would like to automatically create camera groups, optimize cameras and generate the mesh again.
The distance can be the average distance of points seen by each camera, here would be great to use only points over a certain confidence threshold from the model.
The number of groups should be customizable, for example I would like to have some options:
- every group contains at least N cameras
- groups are generated by manual distance thresholds (ex. group 1 from 0 to 0.5m, group 2 from 0.5 to 1m, group 3 from 1 to 2m, etc..)
- automatic grouping based on some clustering algorithm / histogram binning