Subdividing non-rigid 3D objects into quasi-rigid parts

A new method for efficient modelling of image sequences has been proposed by Kappei and Liedtke (1987) using an analysis by synthesis algorithm. It is based on extracting the three dimensional information from a sequence of TV images in order to construct a 3D model of the scene. The method can be used for the efficient transmission of TV images to a receiver, where the image sequence has to be reconstructed from the model. The surface of three dimensional objects is modeled with the help of a mesh of triangles containing small triangular parts of the image. An image of the sequence can be reconstructed from the 3D model object by projecting all the triangular images onto a hypothetical camera target. >