海量散乱点云数据的模糊聚类挖掘方法研究  

Research on Fuzzy Clustering Mining Method for Massive Scattered Point Cloud Data

作　　者：陆兴华[1] 刘文林 吴宏裕 冯飞龙 LU Xing-hua;LIU Wen-lin;WU Hong-yu;FENG Fei-long(Huali College Guangdong University of Technology,Guangzhou 511325,China)

机构地区：[1]广东工业大学华立学院

出　　处：《计算机技术与发展》2019年第11期12-16,共5页Computer Technology and Development

基　　金：2019年“攀登计划”广东大学生科技创新培育专项资金立项项目(pdjh2019b0616)

摘　　要：物联网和云计算环境下海量散乱点云数据挖掘容易受到关联规则项的干扰,数据挖掘的模糊聚类不好。为了提高海量散乱点云数据挖掘能力,提出一种基于支持向量机的大数据分类挖掘技术。采用分段向量量化编码技术进行海量散乱点云数据空间存储结构分析,结合闭频繁项集检测方法进行海量散乱点云数据的信息融合处理,对高维融合数据进行语义特征分析和关联规则特征提取,对提取的海量散乱点云数据的关联规则采用支持向量机分类器进行模式识别,结合尺度分解方法对分类输出的海量散乱点云数据进行降维处理,采用模糊聚类方法实现对海量散乱点云数据的分类挖掘。仿真结果表明,采用该方法进行海量散乱点云数据挖掘的聚类性能较好,数据挖掘的精度较高。In the environment of Internet of things and cloud computing,the mining of massive scattered point cloud data is easily disturbed by association rule items,and the fuzzy clustering of data mining is poor.In order to improve the ability of massive scattered point cloud data mining,we present a classification and mining technique for big data based on support vector machine(SVM).The segmented vector quantization coding technique is used to analyze the spatial storage structure of mass scattered point cloud data,and the information fusion processing of mass scattered point cloud data is carried out by combining the detection method of closed frequent itemsets.Semantic feature analysis and association rule feature extraction are carried out for high-dimensional fusion data.Support vector machine classifier is used to recognize the association rules of massive scattered point cloud data.Based on the scale decomposition method,the dimensionality of the massive scattered point cloud data is reduced,and the classification mining of the massive scattered point cloud data is realized by using the fuzzy clustering method.The simulation shows that the clustering performance of this method is better and the precision of data mining is higher.

关 键 词：海量散乱点云数据 挖掘 模糊聚类 特征提取 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]