基于SVM的点云模型识别技术的研究  

Point clouds model recognize method based on the recognize technology of support vector machine

作　　者：常伟杰[1] 蔡勇[1] 蒋刚[1] 韩晓东[1] 

机构地区：[1]西南科技大学制造科学与工程学院,绵阳621010

出　　处：《机械设计与制造》2009年第9期104-106,共3页Machinery Design & Manufacture

基　　金：国家自然科学基金资助项目(10576027)

摘　　要：提出了一种基于支持向量机统计识别技术的点云曲面重构法,直接从点云中识别出模型并进行模型替代,从而很大程度上减小了数据量并且保证了建模精度。实验证明,此方法具有较好的效果,并有误差小、速度快等优点。An efficient method for point clouds model recognize based on the recognize technology of support vector machine is presented. The point clouds model is recognized from the point sets and replaced by a new model,so the data mount of the point clouds is reduced and the modeling accuracy is guaranteed. Through the experiment,a nice effect is acquired,with the benefits of excellent error accuracy, high efficiency.

关 键 词：支持向量机 点云 模型 识别 

分 类 号：TH16[机械工程—机械制造及自动化] TP301[自动化与计算机技术—计算机系统结构]