zLayer: Simulating depth with extended parallax scrolling

Abstract

This paper describes an image-based primitive, zLayer, for modeling virtual environments. It serves as a faster alternative to traditional modeling and rendering of three-dimensional objects but constrains the viewpoint to translation only. Current image-based approaches interpolate a new frame using frames created from neighbouring viewpoints. In contrast, zLayers are created from a single viewpoint. To simulate camera translation, new frames are interpolated by warping and compositing multiple zLayers. This paper describes the creation process of zLayer from three-dimensional computer models, photographs and hand drawn images. The generation of zLayer from computer model can be automated. Finally, various applications of zLayers are discussed.