A Meshing Scheme for Efficient Hardware Implementation of Butterfly Subdivision Surfaces using Displacement Mapping

A common bottleneck for high performance graphics systems is the transfer of millions of triangles from the CPU to the graphics pipeline. Displacement Mapping is an effective technique for encoding high levels of detail of surface models through the utilization of coarse triangle meshes together with displacement maps. This technique can be employed to reduce the communication requirements between the CPU and the graphics pipeline. In this paper we present a new adaptive subdivision scheme in which the curvature of the mesh together with the displacement map information is considered. We also present a hardware architecture for the implementation of this proposal. This hardware is an addition to the standard pipeline in order to achieve a displacement map rendering pipeline. The hardware architecture proposal is regular, with an efficient structure that minimizes data storage and avoids the wait cycles that would be associated with the multiple data accesses required per subdivision. With this method, high quality images are obtained with low communication requirements as only the coarse mesh has to be sent. Moreover, the proposed data management permits reemployment of the information in such a way that each triangle has to be sent only once from the CPU to the graphics pipeline. Keywords— Adaptive Subdivision, Displacement Mapping, Mesh Storage, Graphics Hardware.