View-Dependent Multi-Resolution Model

摘要

Multi-resolution model is an effective approach to the problem of rendering complex environment. Most previous research work concentrated on generating view-independent multi-resolution model, which has some difficulties when rendering large-volume models. In this paper, we introduce a view-dependent multi-resolution model, which can create appropriate level of detail at run-time. We use the criteria including the view frustum, face orientation, face distance and screen project area as well as the supplemental factors to simplify and restore the models for real-time rendering. In the paper, we also propose some strategies of switching LODs.