数字化人体表面三维重建及显示的有关问题

Several problems about 3D surface reconstruction and display of digital human

目的 :探讨使用切片技术建模的“虚拟人体”图库 ,重建人体三维图像的表面形状。方法 :在目前通用的个人计算机装备上 ,结合计算机的视觉、数据存储、图形学和数字图像处理等技术 ,重建人体表面三维图像。结果 :在大图库基础上建立了小图库 ,减少了数据量 ;通过滤镜除去噪点 ,提取边界 ;确定可见边界点 ,消除冗余边界点 ;构成前视冠状图 ;将静态显示制成动画 ;用实时计算显示 ,解决可控问题。结论 :探讨了变处理大图库大数据量为小图库小数据量 ,建立表面三维要素数据库 ,减少计算机的计算量 ,加速三维重建及显示的过程 ,对未来“虚拟中国人”应用有较宽广的覆盖面。

Objective: To reconstruct three-dimensional surface image of human from image library of 'Virtual Human' which was made up of serial human sections. Methods: According to the computer vision, computer data saving, computer graphics and computer image processing, three-dimensional surface image of human was reconstructed. Results: Little-image-library was build up from big-image-library to reduce calculation; the borderline of specimen in the image was recognized after removing noise by filter; the front view of the specimen corona was established; static image was changed into moving picture; at the same time of computing, moving picture was showed, so that user can control in a more great degree. Conclusions: In the article, changing big-image-library into little-image-library, building the database of 3D surface factors from little-image-library, reducing calculation and accelerating 3D reconstruction and display are proposed and explored. In the future, all this techniques will be applied to Virtual Chinese Human in some ways.