4. Applications

When an object seen by the calibrated camera is known, it is possible to calculate its pose, i.e. its position and orientation in space, from measurements made in the image. This is known as image-based pose calculation. In this context, "knowing an object" means that remarkable visual primitives have their coordinates, expressed in the object's frame of reference, known. Remarkable visual primitives are usually points, segments, circles or other geometric patterns, or even the object's intensities directly. They must be "remarkable" to enable them to be measured in the image, and thus achieve 2D-3D correspondence, i.e. between the 2D image and the 3D object model.