Interactive model reconstruction with user guidance

We have demonstrated how augmented reality can be used to assist view planning during rapid model acquisition. Using the partial model generated from ProFORMA, the viewing sphere is split up into an icosahedron, with scores applied to each face depending on visibility of model triangles and keyframe object orientations. This information is then used to calculate a motion, which is presented to the user in the form of a spatial arrow registered to the object. In this way, the user is guided to collect new keyframes from desirable views, allowing completion and refinement of the model. User guided modelling using ProFORMA enables a complete 3D textured model to be created in around one minute, including sequence collection time.