%0 Journal Article %T 一种改进的DTW算法在人体行为识别中的应用 %A 顾军华 %A 徐俊生 %A 刘洪普 %J 河北工业大学学报 %D 2018 %R 10.14081/j.cnki.hgdxb.2018.04.004 %X 目前针对各维特征之间的相关性对动作识别影响的问题,解决的方法基本是采用欧氏距离作为其相似 度的算法.结合Kinect体感设备提出了一种基于改进的DTW算法的人体行为识别方法,将马氏距离作为相似度 测度引入DTW算法中进行改进.首先,利用Kinect设备采集人体骨骼信息,计算骨骼夹角信息,之后使用改进 的DTW算法进行模板训练和人体行为识别.最后通过实验,对比采用欧氏距离、卡方检验和马氏距离作为相 似度测度时,人体行为识别的准确率.实验表明引入马氏距离的DTW算法在识别正确率方面有所提高.</br>AhumanbehaviorrecognitionmethodbasedonanimprovedDTWalgorithmisproposedbycombiningthe Kinectsensorinthepaper.Tosolvetheproblemthatifthecorrelationbetweeneveryfeaturewillaffectbehaviorrecogni? tion,Euclideandistanceisadoptedintheexistingwayasitssimilarityalgorithm.Mahalanobisdistanceisintroducedin? totheDTWalgorithmassimilaritymeasure.Firstly,weuseKinectsensortocollectthehumanskeletoninformation;then theimprovedDTWalgorithmisadoptedfortemplatetrainingandhumanbehaviorrecognition.Finally,compareaccura? cyrateofhumanbehaviorrecognitionwhenEuropeandistance,Mahalanobisdistanceandchi-squaretestseparately serveasthemeasureofsimilaritydetection.TheexperimentsshowthattheDTWalgorithmintroducedbytheMahalano? bisdistancehasbeenimprovedintherecognitionaccuracyrate. %K 行为识别 %K 动态时间规整 %K 欧氏距离 %K 马氏距离 %K 卡方检验 %K 距离测度< %K /br> %K behaviorrecognition %K dynamictimewarping %K Euclideandistance %K Mahalanobisdistance %K chi-squaretest %K distancemeasure %U http://zrxuebao.hebut.edu.cn//oa/darticle.aspx?type=view&id=201804004