Architectural extensions to 3D texturing units for accelerated volume rendering

摘要

Volume rendering of an image from a volume data set is performed by dividing the image into a plurality of sub-regions and processing the sub-regions one at a time. The sub-regions are processed one at a time by a chip including a triangle unit for creating projections of sub-region on at least one plane cutting the volume data set; a rasterization/3D-texturing unit for texturing each projection received from the triangle unit; and a triangle buffer and compositing unit for compositing each projection textured by the rasterization/texturing unit. After a sub-region has been processed, the chip sends the composited projection to an external frame buffer for display.