基于聚类分析的高维数据异常特征光流检测系统

Optical flow detection system based on clustering analysis for abnormal characteristics of high-dimensional data

针对高维数据异常特征光流检测精度较低的问题,提出一种基于聚类分析的高维数据异常特征光流检测系统,分别介绍了系统硬件中的控制单元模块、电源电路模块、接口模块;在系统的软件设计部分,采用聚类分析法设计异常特征聚类模块,在此基础上设计异常特征光流检测模块,最后给出系统数据库,完成系统软件设计,通过硬件与软件的组合实现了高维数据异常特征的光流检测。为了验证该系统的检测精度,进行对比实验,实验结果表明该系统的检测精度远高于传统高维数据异常特征光流检测系统,证明该系统更适用于高维数据异常特征的光流检测。

Aiming at the low accuracy of high-dimensional data anomaly optical flow detection,a high-dimensional data anomaly characteristic optical flow detection system based on cluster analysis is proposed.The control unit module,power supply circuit module and interface module in the system hardware are introduced respectively.In the software design part,the cluster analysis method is used to design the abnormal feature clustering module.On this basis,the abnormal feature optical flow detection module is designed.Finally,the system database is given,and the system software design is completed.Then the combination of hardware and software realizes the optical flow detection of dimensional data anomaly features.In order to verify the detection accuracy of the system,a comparison experiment is carried out.The experimental results show that the detection accuracy of the system is much higher than that of the traditional high-dimensional data anomaly characteristic optical flow detection system,which proves that the system is more suitable for optical flow detection of high-dimensional data anomaly features.