简介：

简介：A3D扩充了现实航行系统使用实体镜的图象为teleoperated机器人系统被开发。在模仿的模型和实际机器人的录像图象之间的精确匹配能被认识到，它帮助操作员正确地并且可靠地完成遥控任务。系统介绍不同地图翻译转变方法为实体镜的显示器拿视差图象，提供操作符immersive3D经验。同时，动态立体声录像的一个快、精确的登记方法被建议，并且一个虚拟机器人的有效集成和真实立体声景色能被完成。系统的操作错误在不到2.2公里和平均错误被维持的初步的实验表演是0.8547，在x的0.9093和0.6972公里，y，z方向分别地。大量实验象按按钮那样，等等拉抽屉也被进行评估系统的表演。结构的深度信息能是的可行性研究表演很快并且在遥远的环境认出了地点。图象覆盖系统的扩充现实能增加操作精确性并且由于看的直觉的3D减少过程时间。

简介：Single-cameramobile-visioncoordinatemeasurementisoneoftheprimarymethodsof3D-coordinatevisionmeasurement,andcodedtargetplaysanimportantroleinthissystem.Amultifunctionalcodedtargetanditsrecognitionalgorithmisdeveloped,whichcanrealizeautomaticmatchoffeaturepoints,calculationofcamerainitialexteriororientationandspacescalefactorconstraintinmeasurementsystem.Theuniquenessandscalabilityofcodingareguaranteedbytherationalarrangementofcodebits.Therecognitionofcodedtargetsisrealizedbycross-ratioinvariancerestriction,spacecoordinatestransformoffeaturepointsbasedonspacialposeestimationalgorithm,recognitionofcodebitsandcomputationofcodingvalues.Theexperimentresultsdemonstratetheuniquenessofthecodingformandthereliabilityofrecognition.

简介：Arobustsystemforbacklitkeyboardinspectionisrevealed.Thebacklitkeyboardnotonlyhaschangeablediversecolorsbutalsohasthelasermarkingkeys.Thekeysonthekeyboardcanbedividedintoregionsoffunctionkeys,normalkeys,andnumberkeys.However,theremighthavesometypesofdefects:incorrectilluminatingarea,non-uniformilluminationofspecifiedinspectionregion(IR),andincorrectluminanceandintensityofindividualkey.Sincetheilluminationfeaturesofbacklitkeyboardaretoocomplextoinspectforhumaninspectorintheproductionline,anauto-matedinspectionsystemforthebacklitkeyboardisproposedinthispaper.Thesystemwasdesignedintotheoperationmoduleandinspectionmodule.Asetofimageprocessingmethodsweredevelopedforthesedefectsinspection.Someexperimentalresultsdemonstratetherobustnessandeffectivenessoftheproposedsystem.

简介：Thetraditionaleyesensitivityfunctionbasedonphotopicvisionisnotapplicableinthemesopicvisionstate.Themesopicvisionisstudiedbyusingthephotopicandscotopicvisionstatesensitivityfunctions.Andthemodelwhichlinksthemesopicsensitivitywiththephotopicandscotopicstatesisbuilt.Basedonthemodel,theluminousefficacyofmesopicvisioniscalculatedandappliedtothespectrumdistributionofLEDlightsources.TheresultsshowthattheluminousefficacyofacommercialYAGphosphorconvertedwhiteLEDisupto172.7lm/Watmesopicvision,whichis67.2%higherthanthatofphotopicvisionstate.WealsoconcludethattheoptimalspectralpowerdistributionoftheLEDwillgreatlyincreasethemesopicluminousefficacy.

简介：Afeature-constrainedstereomatchingalgorithmforlunarrovernavigationispresentedbasedontheanalysisofthestereovisionsystemandworkingenvironmentsoflunarrover.Infeature-matchingphase,edgepointsareextractedwithwavelettransformandareusedastheprimitivesformatching.Thenthreecriterionsareutilizedinturntoselectthecorrectmatchingpointswiththepyramidalsearchingstrategy.Asaresult,thealgorithmfindscorrespondingpointssuccessfullyforlargenumbersofedgepoints.Area-matchingisaccomplishedundertheconstraintofedge-matchingresults,andthecorrelationisselectedasthecriterion.Experimentalresultswithrealimagesofnaturalterrainindicatethatthealgorithmprovidesdensedisparitymapswithfairlyhighaccuracy.

简介：Vehicleanti-collisiontechniqueisahottopicintheresearchareaofIntelligentTransportSystem.Theresearchonprecedingvehiclesdetectionandthedistancemeasurement,whicharethekeytechniques,makesgreatcontributionstosafe-driving.Thispaperpresentsamethodwhichcanbeusedtodetectprecedingvehiclesandgetthedistancebetweenowncarandthecarahead.Firstly,anadaptivethresholdmethodisusedtogetshadowfeature,andashadow!areamergingapproachisusedtodealwiththedistortionoftheshadowborder.Regionofinterest(ROI)isobtainedusingshadowfeature.ThenintheROI,symmetryfeatureisanalyzedtoverifywhethertherearevehiclesandtolocatethevehicles.Finally,usingmonocularvisiondistancemeasurementbasedoncamerainteriorparametersandgeometricalreasoning,wegetthedistancebetweenowncarandtheprecedingone.Experimentalresultsshowthattheproposedmethodcandetecttheprecedingvehicleeffectivelyandgetthedistancebetweenvehiclesaccurately.

简介：Anewrnultiscaleedgedetectionmethodispresented,whichisbasedonaneffectiveedgemeasure.Theeffectiveedgemeasure,usedtoadaptivelyadjustthescalesofwavelettransform,isdefinedusingthenovelfeaturesofimageedgeobtainedfromhumanbeingvisioncharacteristics.Finally,twoexperimentsshowthattheproposedalgorithmappearstoworkwell.

简介：Thisstudyanalyzeslivefacialvideosforrecognizingnonverballearning-relatedfacialmovementsandheadposestodiscoverthelearningstatusofstudents.First,coloranddepthfacialvideoscapturedbyaKinectareanalyzedforfacetrackingusingathree-dimensional(3D)activeappearancemodel(AAM).Second,thefacialfeaturevectorsequencesareusedtotrainhiddenMarkovmodels(HMMs)torecognizesevenlearning-relatedfacialmovements(smile,blink,frown,shake,nod,yawn,andtalk).Thefinalstageinvolvestheanalysisofthefacialmovementvectorsequencetoevaluatethreestatusscores(understanding,interaction,andconsciousness),eachrepresentsthelearningstatusofastudentandishelpfultobothteachersandstudentsforimprovingteachingandlearning.Fiveteachingactivitiesdemonstratethattheproposedlearningstatusanalysissystempromotestheinterpersonalcommunicationbetweenteachersandstudents.

简介：Inthispaper,wepresentarobotvisionbasedsystemforcoordinatemeasurementoffeaturepointsonlargescaleautomobileparts.Oursystemconsistsofanindustrial6-DOFrobotmountedwithaCCDcameraandaPC.Thesystemcontrolstherobotintotheareaoffeaturepoints.Theimagesofmeasuringfeaturepointsareacquiredbythecameramountedontherobot.3Dpositionsofthefeaturepointsareobtainedfromamodelbasedposeestimationthatappliestotheimages.Themeasuredpositionsofallfeaturepointsarethentransformedtothereferencecoordinateoffeaturepointswhosepositionsareobtainedfromthecoordinatemeasuringmachine(CMM).Finally,thepoint-to-pointdistancesbetweenthemeasuredfeaturepointsandthereferencefeaturepointsarecalculatedandreported.Theresultsshowthattherootmeansquareerror(RMSE)ofmeasurevaluesobtainedbyoursystemislessthan0.5mm.Oursystemisadequateforautomobileassemblyandcanperformfasterthanconventionalmethods.

简介：

简介：

简介：有了星友提供的故障原因，笔者下手自然就不会再通电测试了，而是直接拆机检查。

简介：Thedisadvantageremainingin3Dscanningsystemanditsreasonsarediscussed.Anewhost-and-slavestructurewithhighspeedimageacquisitionandprocessingsystemisproposedtoquickentheimageprocessingandimprovetheperformanceof3Dscanningsystem.

简介：今年NAB上，人们更多关注的是3D电视广播的可能性及对技术的探讨。像Quantel．Avid这样的顶级非编系统提供商，就将3D工作流程集成到了制播系统中。特别是重返NAB的Quantel．凭借其DI与3D方面优势，及在《阿凡达》等3D电影中的参与，在展会上浓墨重彩的推广是理所当然的。

简介：北京2010年8月4日电，美通社亚洲/--2010年初，3D影视巨制《阿凡达》所点燃的第三次3D产业发展浪潮曾经一度如火如荼，然而，这场热潮现在正面临“倒退”的危机，良莠不齐的3D内容及3D产品充斥市场，“劣币驱逐良币”，伪3D产品严重危害消费者健康和3D产业健康成长。

简介：2010年,3D潮流可谓锐不可挡,以主动快门式3D技术为代表的眼镜式3D电视大规模登陆市场,市场销售呈现开门红,预计全年全球销量将突破300万台,因此2010年可以称得上真正的"3D电视发展的开元之年"。眼镜的束缚似乎没有阻挡消费者对3D电视的欣赏热情,但冷静之余,我们却不得不面对眼镜式3D电视可能会对视力造成的不良影响,以及长时间观看所带来的头晕、恶心等不良感受,而且对近视眼患者因佩戴"双重"眼镜所带来的痛苦更是不可言状,因此,我们不得不承认,目前的眼镜式3D电视只能作为3D电视技术发展的过渡产品,而最终必然被成熟的裸眼3D电视技术所取代,裸眼3D显示将是未来3D电视技术发展的必然归宿。

简介：<正>3D打印机打印了枪支!打印了飞机!打印了房子!还打印了人体器官……几乎每天都有关于3D打印的神奇报道,令人兴奋不已,又令人充满无限的好奇。这是什么技术?3D打印(3Dprinting),是一种以数字模型文件为基础,运用粉末状金属或塑料等可粘合材料,通过逐层打印的方式来构造物体的技术。它也是快速成型技术的一种。3D打印技术完全不同于传统加工技术,具有仿真性强、速度快、价格便宜、高易用性等优

简介：音频A/D转换芯片的输出增益和失真度是无法直接测试的,需要对输出的数字信号进行解析还原成模拟信号方可以进行测试。有些A/D产品输出的数字信号可以直接进行解析还原成模拟信号,而有些则需要先对输出的数字信号进行处理。以某A/D转换芯片为例,首先利用了六反向施密特触发器74HC14和8位串入并出移位寄存器74HC164对输出信号时序进行转换,然后采用stm32f103对时序转换后的数字信号进行解析,最后将解析后的数据传送至上位机,最终实现了对该A/D转换芯片输出增益及失真度测试的目的。

简介：Intersil公司日前宣布它已签署了一项收购D2Audio公司的最终协议，后者是一家专门致力于为消费电子、汽车以及专业音频产品等应用领域提供数字音频功率放大器设计的创新领先厂商。至此，Intersil公司成为全球独家拥有智能数字音频功率放大器技术的厂商。