Volume　rendering　could　display　the　image　in　an　integrated　distribution　of　a　variety　of　substances.　It　is　convenient　to　view　the　body　data　for　a　doctor.　Volume　rendering　should　traverse　each　voxel　of　the　data,　causing　large　calculation　and　slow　image　generation,　and　can　not　change　the　external　light　and　perspective.　It　is　need　to　start　the　drawing　again　when　parameters　have　changed.　The　complex　process　lead　the　slow　speed　of　rebuilt.　In　this　paper,　volume　rendering　algorithm　based　on　segmentation　in　three-dimensional　reconstruction　is　present,　including　the　segmentation　of　CT　data,　the　normal　vector　of　the　volume　element,　and　the　generation　of　the　final　image.　This　method　greatly　accelerated　the　speed　of　rendering　image,　and　improves　image　display　quality.　Also,　it　can　well　reflect　the　global　status　of　three-dimensional　data.　©　2012　IEEE.