双线激光轮廓测量中点云数据的平滑方法  

Smoothing Method of Point Cloudy Data for Double Line Laser Profile Measuring

作　　者：张国敏[1] 姚剑敏[1] 郑建用[1] 

机构地区：[1]福州大学物理与信息工程学院,福建福州350108

出　　处：《计算机与现代化》2013年第10期65-68,共4页Computer and Modernization

基　　金：国家自然科学基金资助项目(61101169);国家863计划项目(2012AA03A301);国家大学生创新创业训练计划项目(201210386024)

摘　　要：基于双线激光的三维扫描系统在点云数据拼接中易导致轮廓粗糙。本文提出一种基于横向离散点拟合的表面平滑方法。该方法首先采用平均算法对两个截面上的点进行平滑,其次通过选取每帧激光条纹上间隔较远的两个点进行曲线拟合,由这两个点的移动来带动整条激光条纹的微调。实验结果表明,根据所提出的方法调整后的离散点重构出的三维模型,其表面的平滑性得到了进一步的改善,重构模型表面形状与实物的表面形状更加接近。Double line laser 3D scanning system is easy to cause the rough outlines in the point cloud data fusing. This paper pro- poses a smooth surface based on transverse fitting of discrete points. The average algorithm is applied to smooth the two section points on the curve fitting and followed by se]ecting two points per - frame on the interval of laser stripe further. It leads the entire laser stripes trim by moving these two points. According to the proposed method of discrete points to reconstruct the 3D model, the experimental results show that the surface smoothness is improved and the reconstruction surface shape model is more close to real surface shape.

关 键 词：离散点 曲线拟合 三维模型 三维重构 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]