共聚焦三维数据表面重建的一种反走样方法  

Anti-aliasing technique for surface reconstruction of confocal data

作　　者：薛斌党[1] 姜志国[1] 周孝宽[1] 

机构地区：[1]北京航空航天大学宇航学院,北京100083

出　　处：《北京航空航天大学学报》2005年第10期1054-1057,共4页Journal of Beijing University of Aeronautics and Astronautics

基　　金：国家自然科学基金资助项目(60372017);Motic北航研发中心资助项目

摘　　要：用边界体素的集合表示物体表面,将边界体素作为点投影到屏幕上形成光照图像时,可能出现空洞和图像走样.改进了体素表面算法:用剖分立方体算法得到物体的边界体素集,然后用边界体素的中心点构造投影表面点,用脚印函数计算表面点对显示图像平面上像素的影响范围,以此值作为该像素的权值,形成最后的图像.这样,边界体素不是只投影到一个点上,而是有一个投影区域,从而避免空洞现象,并减轻走样现象.实验证明,该方法具有较快的表面重建速度,消除了空洞,改善了显示效果.There might exists missing holes and aliasing on the projected image by way of representing object surface with boundary voxels and projecting boundary voxels to the screen as points to render the shading image of the object. An anti-aliasing technique for surface reconstruction is proposed. Firstly, the boundary voxel sets is generated by dividing cubes; Secondly, the values of the center point of the boundary voxel sets is obtained to form the surface point sets ; Then, a footprint function is chosen to calculate the effect region of the surface point on the pixel on the screen, and the result value is used as the weight of the pixel. This processing is just like splatting, so the missing holes is eliminated and the display quality is improved. Experimental results show that both the surface reconstruction and the display performance of the technique are satisfactory.

关 键 词：共聚焦成像 表面重建 反走样 剖分立方体算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]