Publisher Summary This chapter describes a system for the analysis and synthesis of a sequence of images. The system relies on Motion Analysis and Learning techniques. It is able to extract, from an image sequence, some control parameters describing the motion contained in the sequence, and is able to reconstruct the original images of the sequence by using only such control parameters and a small set of the input images. The realized system has many practical applications in the domains of (1) computer graphics, (2) data compression, (3) teleoperation and object recognition system, and (4) virtual reality. The information to be extracted is application dependent.
