Image-based modeling using viewpoint entropy

Abstract

We present a new method to automatically determine the correct camera placement positions in order to obtain a minimal set of views for Image-Based Rendering. The viewpoints should cover all visible polygons with an adequate quality, so that we sample the polygons at enough rate. This allows to avoid the excessive redundancy of the data existing in several other approaches. The localisation of interesting viewpoints is performed with the aid on an information theory-based measure, dubbed viewpoint entropy. This measure can be used to determine the amount of information seen from a viewpoint. We have also developed a greedy algorithm that aims to minimise the number of images need to represent a scene. In contrast to other approaches, our system uses a special preprocess for textures to avoid artifacts appearing in partially occluded textured polygons. Therefore no visible detail of these images is lost.