Conversion of polygonal surfaces to displaced subdivision surfaces

Displaced subdivision [9] puts forth a number of attractive features for editing, geometry compression, animation, scalability, and adaptive rendering of polygonal models. In this representation, a detailed surface model is defined as a scalar-valued displacement map over a smooth domain surface. The construction of the smooth domain surface from a polygonal model is a challenging task in the conversion process. We propose a new method for defining the smooth domain surface based on root 3-subdivision scheme and a linear time optimization technique. At some fixed level of detail, the vertex and triangle complexity of the displaced surface generated by the proposed method is far less and the magnitude of the offset values defining the displacement map are smaller and so it results in higher compression ratios and better transmission speed. The proposed algorithm creates surfaces of better quality, is computationally more efficient and occupies less memory as compared to the original algorithm [9]