Decompression of Packed Vertices on Graphics Hardware

People: Jonathan Bilodeau, Budirijanto Purnomo, Jonathan D. Cohen, and Subodh Kumar.
Description: We present a vertex compression technique suitable for efficient decompression on graphics hardware. Vertex data maybe transmitted to and stored in the video memory in the compressed form, and the vertices are decompressed on the fly using a vertex program. We present algorithms for automatically quantizing vertex attribute data using intrinsic properties, triangle clustering to achieve further lossless compression on top of the quantization, and packing quantized attributes into hardware-compatible vector variables. We also describes the vertex program capable of efficiently decompressing these packed vertex attributes at rendering time. Our algorithm works well within the constraints of current graphics hardware and provides a useful tool for optimizing space and bandwidth constraints of interactive graphics applications.
Papers/TechReport: Decompression of Packed Vertices on Graphics Hardware
Jonathan Biladeau, Budirijanto Purnomo, Jonathan D. Cohen and Subodh Kumar
Johns Hopkins Tech Report [JHU-CS-GL04-2]

For more information,  Jonathan Cohen (cohen@cs.jhu.edu , Michael Kazhdan (misha@cs.jhu.eduFaculty
                                    Yuan Chen (cheny@cs.jhu.edu)
Webmaster