We consider a scene, containing many objects moving with constant v elocity along straight line paths, seen from three reference viewpoints at three di erent times. The scene may e v en consist only of moving objects with no static features. We wish to create a new image sequence showing the scene from any arbitrary viewing position. We make use of a newly discovered tool, the dual Htensor" 1 , that connects together three views of a coplanar con guration of unlabeled static and moving points.

more »

... sing the dual Htensor we show h o w to factor out the viewing transformation and create a new sequence in which the static features are stable and the moving features are synthesized as if moving with constant v elocity. The constant v elocity is achieved by i n troducing a 1D collineation that transforms the positions of the objects in the three views into their positions at new times. The support of this research b y the O ce of Naval Research under contract N00014-95-1-0521, is gratefully acknowledged, as is the help of Sara Larson in preparing this paper.