76V制衣技术2023年第7期针织工业老年男性腰臀部体型分类及预测模型建立曹晓梦1,2,王春茹1,罗斯祺1,钟安华1（1.武汉纺织大学服装学院，湖北武汉430073；2.江西服装学院完服装工程学院，江西南昌330201）摘要：体型分类参数是服装智能化生产的关键要素，是制衣技术精细化的重要前提，而体型模型的建立与预测是当前服装行业的重要研究课题。为分析目前老年男性函待解决的裤装适体性问题，通过三维人体扫描仪获得129个60~75岁的老年男性人体样本数据，分别通过形态分析和数值分析说明了老年男性腰臀部和青年男性的差异。进一步确定最佳聚类数,并利用K-means聚类法将老年男性腰臀部划分为4类。同时构建基于BP神经网络的体型预测模型，测试集总体识别率达到93.75%。结果显示，该方法可有效区分老年人体腰臀部形态差异，为制作合体性更高的老年男性裤装结构奠定了理论基础。关键词：老年男性；腰臀部体型；体型分类；预测模型；神经网络中图分类号：TS941.17文献标志码：A文章编号：1000-4033(2023)07-0076-05lassificationandPredictionModeloftheWaistandHipTypesoftheElderlCaoXiaomengl?,WangChunru',LuoSiqi',ZhongAnhua'(1.SchoolofFashion,WuhanTextileUniversity,Wuhan,Hubei430073,China;2.SchoolofFashionEngineering,JiangxiInstituteofFashionTechnology,Nanchang,Jiangxi330201,China)Abstract:Bodyshapeclassificationparametersareakeyelementofintelligentclothingproductionandtheimportantpremiseoffinegarmenttechnology,andtheestablishmentofbodyshapemodelsisanimportantresearchtopicinthecurrentapparelindustry.Inordertoanalyzetheproblemoftrousersuitabilityfortheelderlytobesolvedurgently,129humansamplesof60~75yearsoldmenwereobtainedthroughathree-dimensionalbodyscanner.Throughmorphologicalanalysisandnumericalanalysis,respectively,thedifferencebetweenthewaistandhipsoftheelderlyandtheyoungwasexplained,andthesignificanceoftheresearchwasconfirmed.Furtheritdeterminedtheoptimalnumberofclusters,andusedtheK-meansclusteringmethodtodividethewaistandhipsoftheelderlyinto4categories.Onthisbasis,abodyshapepredictionmodelbasedonBPneuralnetworkwasconstructed,andtheoverallrecognitionrateofthetestsetreached93.75%.Theresultsshowthatthismethodcaneffectivelydistinguishthedifferenceinthes...
