基于结构光的三维点云重建方法研究  

Research of three-dimensional point cloud reconstruction based on structured light

作　　者：孙庆科 何云涛[1] 陈瑞强[1] 江月松[1] Sun Qingke;He Yuntao;Chen Ruiqiang;Jiang Yuesong(School of Electronic Information Engineering, Beihang University, Beijing 100191 , China)

机构地区：[1]北京航空航天大学电子信息工程学院,北京100191

出　　处：《计算机应用研究》2016年第9期2873-2877,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61101154)

摘　　要：结构光三维成像是近年来三维成像领域的研究热点之一。整体设计及实现了条纹结构光三维重构系统,并重点研究了点云生成三角网格方法。该系统使用整体阈值与局部滑动阈值相结合的方法提取到条纹中心特征点,以像素索引值为中间媒介进行编码值插值计算,并利用像素索引值为媒介对点云进行三角网格化处理。利用像素索引值的方法简化了点云插值和点云生成三角网格的处理过程,并且能够精确得到每个点上的颜色值并进行颜色渲染。最后利用提出的方法对石膏模型和实际人脸面部进行了三维测量和重建,并分析了该方法测量的精度。结果表明提出的方法达到了实验精度的要求并取得了非常好的三维重构效果。Structured light three-dimensional imaging is one of the hotspots in recent years in the field of three-dimensional imaging. This paper overall designed and implemented a three-dimensional reconstruction system of stripe structured light,and focused on the method of triangle mesh generation of point cloud. It used combination of the global threshold and the local sliding threshold methods to extract the feature points in the strip center. By using the pixel index values of feature point for intermediary,it interpolated encode values and point cloud generated triangular mesh. It made the point cloud interpolation and the point cloud triangulation process greatly simplified by using the pixel index values of feature point for intermediary,and it was able to accurately get the exact color value of each point and do color rendering. Finally,it took the surface of a plaster model and a human face as examples to be measured and reconstructed using the proposed method,and analyzed the accuracy of the measurement methods. This results show that the proposed method achieves the accuracy of test requirements and very good performances.

关 键 词：成像系统 三角网格划分 像素索引值 特征点提取 点云插值 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]