资源描述:
《基于多模态眼动识别的乒乓球运动员动作预测.pdf》由会员上传分享,免费在线阅读,更多相关内容在行业资料-天天文库。
1、科技通报第31卷第6期Vol.31No.62015年6月BULLETINOFSCIENCEANDTECHNOLOGYJun.2015基于多模态眼动识别的乒乓球运动员动作预测蔡冠蓝(郑州航空工业管理学院体育教学部,郑州450015)摘要:眼动的速度最高可达到600°/s,具有比人手更灵敏的反应,因此有更好的实时性和应激性,通过眼动识别可以准确反映人体的动作特征,实现动作的判断和预测。采用眼动识别的图像处理算法实现对乒乓球运动员的动作预测,提高攻球效能,改进乒乓球攻击和防御的目的性。传统方法中对乒乓球运动员的
2、眼动识别算法采用边缘特征融合算法,对运动员的动作变换跟踪性能不好,提出一种基于多模态融合眼动识别的乒乓球运动员动作预测算法。提取边缘特征,进行虹膜定位设计,统计搜索区域灰度直方图分布,建立虹膜颜色与边缘联合特征模板,通过多模态融合眼动时变分析,采用从粗到细的处理方法,在减少匹配对应项的同时进行动作预测相关系数匹配,预测跟踪目标中心,实现动作预测,仿真结果表明,算法对运动员的动作预测准确度高,实现对乒乓球运动员动作的实时跟踪和识别。关键词:乒乓球;图像处理;眼动识别;动作预测中图分类号:TP391文献标识码
3、:A文章编号:1001-7119(2015)06-0043-03PredictionofTableTennisPlayersActionBasedonMultimodalEyeRecognitionCaiGuanlan(DepartmentofManagemengtPhysicalEducation,ZhengzhouInstituteofAeronauticalIndustry,Zhengzhou450015,China)Abstract:Eyemovementspeedcanreachamaximum
4、of600degrees/s,ismoreresponsivethanahumanhand,real-timeandstresssohadbetter,witheyerecognitioncanaccuratelyreflectthemotioncharacteristicsofthehumanbody,thereal⁃izationofjudgmentandpredictionofmovement.Usingimageprocessingalgorithmstoachieveeyerecognition
5、oftableten⁃nisathletesactionprediction,improvetheefficiencyofimprovedobjectivetoattacktheball,tabletennisattackandde⁃fense.Thetraditionalmethodofeyemovementsintherecognitionalgorithmoftabletennisathletesusingtheedgefeaturefusionalgorithm,motiontrackingper
6、formanceofathletestransformationisnotgood,putforwardapredictionalgorithmofmultimodalfusioneyerecognitionbasedontabletennisathletesinaction.Theedgefeatureofirislocalizationdesign,statis⁃ticalsearchregionalgrayhistogramdistribution,theestablishmentofajointf
7、eatureofiriscolorandedgetemplate,throughmultimodalfusioneyetime-varying,usingtheprocessingmethodfromcoarsetofine,inreducingthematchingitematthesametimeshouldbeactionpredictioncorrelationcoefficientmatching,predictionandtrackingtargetcenter,therealizationo
8、fmotionpredictionsimulationresultsshowthatthealgorithm,theactionofathleteshighpredictionaccuracy,real-timetrackingandrecognitionoftabletennisathletesinaction.Keywords:tabletennis;imageprocessing;eyemovementrecogniti