Here we introduce the concept of a 3D video sprite. We outline an approach for their automatic creation and manipulation. 3D sprites consists of a number of synchronous video streams and mark up information that allows virtual viewpoints with respect to a live action video to be rendered and combined with traditional 3D virtual environments. As such they constitute a form of image based rendering. Our generation method uses trinocular stereo matching to provide a morphing channel. The matching of corresponding locations in the images is edge-based and relies on the extraction of suitable epipolar geometries. The resulting edge correspondences are then interpolated and used in an efficient morphing algorithm that operates one scan-line at a time to perform image synthesis. Easily-obtained viewing geometry is used to combine 3D video sprites with computer generated 3D environments. 21 Pages