数据简化中的边界提取方法  

Research of boundary extraction method in data reduction

作　　者：赵伟玲[1] 

机构地区：[1]黑龙江科技大学现代制造工程中心,哈尔滨150022

出　　处：《黑龙江科技学院学报》2013年第5期469-471,490,共4页Journal of Heilongjiang Institute of Science and Technology

基　　金：国家自然科学基金项目(51075128);国家科技重大专项项目(2010ZX04016-012);博士后研究人员落户黑龙江科研启动资助金项目(LBH-Q12019)

摘　　要：针对三维点云数据简化过程中边界特征容易丢失问题,研究了一种点云边界特征提取方法。首先,对点的k近邻进行查找,并进行点的球拟合计算,得到拟合球的半径、点的曲率、点到球心距离。其次,通过数据点周围点的分布均匀性、自适应调节参数公式中的阈值,可以达到检测边界特征的目的。由不同模型的实验数据表明,该算法提取的边界满足了后期数据简化所需。Aimed at preventing boundary feature loss more likely to occur in the process of 3D data reduction,this paper highlights a boundary feature extraction method of point cloud.The method involves firstly seeking the k close neighbors of the sampling point,and obtaining the ball fitting radius,the point curvature,and the distance between the point and the center of the sphere using the ball-fitting method,while getting,and then achieving the boundary feature detection,depending on the distributed uniformity of the closest neighbors of point and self-adaptive adjustment of the threshold value of the parameter equation.The experimental data from the different models verify the ability of the algorithm to give the boundary adequate for the follow-up data reduction.

关 键 词：逆向工程 点云数据 曲率 边界提取 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]