Rigs
A rig is a struture of known dimensions. Its purpose is to locate the model in a scene and is usually a regular shaped object that is below or surrounds the model.
Rigs are not needed when the user supplies the rotation and orientation of the camera for each image. This however, requires an extremely high level of precision.
A good rig provides reference points at the boundary of the object. It is better to have a rig too big, than too small.
The following screen shot shows a typical rig.
What makes a good rig
A good rig has the following attributes :
- A known and stable structure. Lego for example has blocks that are all the same size and made to a high tolerance. The rig needs to be consistently the same shape. Lego can curl at the corners and needs to be pressed down around the whole structure to ensure there are no gaps.
- Readily identifyable points. The lego example above uses alternating coloured blocks. This makes the boundary between blocks easier to find.
- Easy to build and optionally adjust. Lego, for example is easy to adjust. Larger rigs can’t do this but can still be adjusted. The concrete floor of a garage can have several crosses marked on the floor that form the corners of a virtual box.
- Rigs need to be viewable from all required angles. This is easy for small rigs, but larger ones such as digitising a whole car should be designed to accommodate images being taken from all required angles.