一种基于HSI色度空间的磨粒成分分析法  

Debris Analysis with PCA in an HSI Color Model

作　　者：何晓昀[1] 吕植勇[2] 李大光[1] 陈涛[1] 王虹[1] 严新平[2] 

机构地区：[1]武汉理工大学信息工程学院,湖北武汉430070 [2]武汉理工大学可靠性实验室,湖北武汉430070

出　　处：《武汉理工大学学报（信息与管理工程版）》2005年第5期1-4,共4页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：国家自然科学基金资助项目(50275111)

摘　　要：介绍了HSI模型的基本原理,并提出使用HSI模型分析磨粒颜色分布。通过比较不同元素构成的磨粒图片在HSI模型参数上面的颜色分布,来判别该磨粒的元素构成,并且使用主成分分析法对磨粒颜色参数进行了化简,从而可以直观地对磨粒进行分类。实验证明,该方法在铁谱磨粒识别方面效果良好。The principle of an HSI model is introduced. The color features of wear debris are specified by an HSI color model. Each particle is characterized by a set of statistical feature debris from the distribution of an HSI color model. The initial feature set is optimized by principal component analysis. A multidimensional scaling procedure is used for defining the classification plane. The selected features help distinguish copper - base alloy, iron and aluminum particles. The examination result proves that this method can classify the three metallic particles well.

关 键 词：HSI颜色模型 磨粒 主成分分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]