基于压力传感器的乘员体征识别OccupantTypeIdentificationBasedonPresssureSensors摘要车辆乘员类型等体征信息的准确识别是智能安全气囊系统实现的前提条件之一，其识别算法的实时性和准确性对于在碰撞发生时安全气囊的起爆时间和充气强度的控制起着决定性作用。目前国内外研究大多是采用布置在坐椅上的压力传感器检测乘员体重，并以此进行不同体型的乘员类型分类；利用超声波雷达或计算机视觉技术检测乘员与仪表板之间的距离，并以此进行正常或危险坐姿的分类。本文通过对不同类型乘员在不同坐姿下的体压分布信息的精确分析，提出了单纯依靠压力分布传感器，基于乘员类型及坐姿导致的体压动态变化信息进行乘员类型识别的乘员体征识别算法.首先针对乘员体压分布的特点，建立了依据体压敏感点分析的识别特征；其次，依据体压敏感点特征分析，提取了基于体压分布信息的不同类型及坐姿下乘员的主特征描述；最后，利用支持向量机的自学习功能，通过对试验采集到的大量的不同体征乘员特征样本的学习，建立了基于支持向量机的乘员类型模式分类器，实现了对不同类型乘员的有效识别，从而为后续的智能安全气囊的开发探索了一条现实可行且成本低廉的乘员类型的识别方法。关键词：乘员体征信息;智能安全气囊;体压分布;乘员类型;支持向量机ABSTRACTVehicleoccupanttypespositioninformationaccuratelyidentifysuchsignsisintelligenceairbagsystemtorealizeoneofthepremisecondition,therecognitionalgorithmforreal-timeandaccuracyinthecollisionoccurredairbagdetonatortimeandinflatableintensitycontroldecisive.Bothathomeandabroad,mostlyinresearchbyarrangementofpressuresensordetectionchair,andbasedonwhichthecrewweightofdifferenttypecrewtypeclassification;Usingultrasonicdetectingradarorcomputervisiontechnologywiththedistancebetweenthecrewdashboard,andbasedonwhichthenormalordangerouspostureclassification.Basedontheresearchwithfawtechnologycenterprojectofdifferenttypes,throughthecrewindifferentposturebodypressuredistributionofaccurateinformationareanalyzed,andtheonlydependuponthepressuredistribution,basedontheoccupanttypeandsensortosittingbodypressuredynamicchangeinformationthecrewtypesandpostureidentificationsignsrecognitionalgorithms,firstcrewforoccupantbodypressuredistributioncharacteristic,establishesthebodypressuresensitiveanalysisbasedonrecognitioncharacteristic;Secondly,basedonanalysisofcharacteristicsofthebodypressuresensitive,extractionbasedonbodypressuredistributionindifferenttypesofinformationandundertheLordsittingdescription;crewFinally,usingsupportvectormachine(SVM),throughtheself-learningfunctiontotestthecollectedalotofdifferentsignsoflearning,crewfeaturessamplesbasedonsupportvectormachine(SVM)wasestablishedtheoccupanttypeandposturemodeclassifier,realizethedifferenttypesandsittingposition,andtheeffectiverecognitioncrewforthesubsequentdevelopmentoftheintelligentairbagexploringarealisticandlow-costoccupanttypeandposturerecognitionmethod.KeyWords:OccupantSignsInformation;IntelligentAirbags;BodyPressureDistribution;OccupantType;SupportVectorMachine目录1绪论..............................................................11.1乘员体征识别技术的背景..........................................11.2乘员体征识别技术的研究现状和未来的发展..........................11.2.1乘员体征识别技术的研究现状.................................11.2.2乘员体征识别技术的发展.....................................41.3乘员体征识别技术的研究目的......................................51.4本文的主要研究内容..............................................52乘员体征识别技术的理论基础和支持向量机方法............................62.1乘...
