Skip to Main Content
This paper proposes a novel 3D piecewise planar reconstruction algorithm, which utilizes the statistical error between a particular frame and its prediction to refine a coarse 3D piecewise planar representation. The algorithm aims utilization of 3D scene geometry to remove the visual redundancy between frame pairs in any predictive coding scheme. This approach associates the rate increase with the quality of representation for determining an efficient description for a given budget. The preliminary experiments on synthetic and real data indicate the validity of the rate-distortion based approach.