基于SVM和KNR的体型分类算法  

Body Shapes Classification Algorithm Based on SVM and KNR

作　　者：先诗亮 刘本永 XIAN Shiliang;LIU Benyong(College of Big Data and Information Engineering,Guizhou University,Guiyang 550025,China;Intelligent Information Processing Lab,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州大学大数据与信息工程学院,贵阳550025 [2]贵州大学智能信息处理研究所,贵阳550025

出　　处：《智能计算机与应用》2020年第11期113-116,共4页Intelligent Computer and Applications

基　　金：国家自然科学基金(60862003)。

摘　　要：本文提出一种基于人体宽高比的体型分类算法。首先提取人体主躯干(肩宽、体长)数据,减小衣着、手臂对体型的影响,然后以主躯干的宽高比为特征,尝试用KNR(核非线性回归)实现体型分类,并与SVM(支持向量机)的性能相比较。实验结果表明,KNR总体取得了较好的分类效果。This paper proposes a body type classification algorithm based on the aspect ratio of human body.Firstly,the main body(shoulder width,body length) data were extracted to reduce the influence of clothing and arms on body type.Then,the body type was classified by KNR(nuclear nonlinear regression) and compared with SVM(support vector machine) based on the aspect ratio of main body.The experimental results show that the KNR has achieved a good classification effect.

关 键 词：体型识别 主躯干 宽高比 支持向量机(SVM) 核非线性回归(KNR) 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]