資源描述:
《【碩士論文】農(nóng)業(yè)機(jī)器人導(dǎo)航線提取算法的研究與實(shí)現(xiàn).pdf》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在教育資源-天天文庫(kù)。
1、Ylto6553分類號(hào).藏密蛆:鳊號(hào).惦2啦s5鬻碩士學(xué)位論文農(nóng)業(yè)帆器人導(dǎo)航線提取算法的研究與實(shí)現(xiàn)學(xué)位中請(qǐng)人:£整墨導(dǎo)師赴#硅職稱剜紅梅副教授專業(yè)£你:墮笪量焦墨墨些中文摘要農(nóng)業(yè)機(jī)器人導(dǎo)航線提取算法的研究與實(shí)現(xiàn)專業(yè):通信與信息系統(tǒng)碩士生:尹建媛指導(dǎo)教師:劉紅梅副教授摘要基于圖像處理技術(shù)的視覺(jué)導(dǎo)航成為農(nóng)用AGV(AutomaticGuidedVehicle)的重要發(fā)展方向,近年來(lái)得到了廣泛的關(guān)注。視覺(jué)導(dǎo)航是通過(guò)圖像處理獲耿農(nóng)用AGV的導(dǎo)航基準(zhǔn)線來(lái)描述行進(jìn)路線,現(xiàn)有方法多是提取一些標(biāo)志點(diǎn)后再由這些點(diǎn)得到相應(yīng)的直線作
2、為導(dǎo)航基準(zhǔn)線。由于農(nóng)田環(huán)境復(fù)雜,農(nóng)用AGV行進(jìn)中多次采集圖像所得到的直線肯定無(wú)法保持一致,即行進(jìn)過(guò)程中不斷得到的導(dǎo)航基準(zhǔn)線組合起來(lái)所描述的行進(jìn)路線是一條折線,因此農(nóng)用AGV的導(dǎo)航基準(zhǔn)線采用曲線更為合理。論文主要研究了曲線導(dǎo)航線的提取算法和優(yōu)化,主要內(nèi)容如下:l,圖像的獲墩和處理。圖像處理中最為關(guān)鍵的是圖像分割。本文采用超綠分割特征,閾值分割法獲得二值化圖像,然后運(yùn)用數(shù)學(xué)形態(tài)學(xué)組合算法進(jìn)行圖像處理,使分割效果更好。2.導(dǎo)航特征點(diǎn)的提取。提取直線導(dǎo)航基準(zhǔn)線要求導(dǎo)航特征點(diǎn)盡量在直線附近,而曲線導(dǎo)航基準(zhǔn)線曲率不確定,需
3、要新的提取標(biāo)準(zhǔn)。本文以行進(jìn)路線彎曲方向?yàn)榍€特征點(diǎn)。3.曲線導(dǎo)航線的擬合。將導(dǎo)航特征點(diǎn)以合適的曲線擬合算法得到導(dǎo)航基準(zhǔn)線。4.導(dǎo)航基準(zhǔn)線提取算法的優(yōu)化。定義基于曲線導(dǎo)航基準(zhǔn)線的行走誤差,利用模糊融合技術(shù)對(duì)提取算法進(jìn)行評(píng)價(jià)和優(yōu)化。關(guān)鍵詞:圖像分割,數(shù)學(xué)形態(tài)學(xué),特征點(diǎn)提取,導(dǎo)航線,模糊決策英文摘要ResearchandImplementationofRouteExtractionAlgorithmofAgriculturalRobotMajor:TelecommunicationandInformationSyst
4、emsName:YinJianyuanSupervisor:ViceProfessorLiuHongmeiABSTRACTVisualnavigationbasedonimage—processingisoneofimportantdevelopingtrendinagriculturalAGV,whichhascausedwideattention.Visualnavigationdescribethemarchingroutinebasedonnavigationdatumlineofagricultura
5、lAGV,whichisacquiredbyimage—processing.Presently,mostmethodsareextractingsomemarkingpoints,thenconnectthemasnavigationdatumline.Becauseofthecomplicatedenvironmentinfarmland,theconnectedlinescannotkeepconsistentbutabendingline.AsaresuR,curveismoresuitablefora
6、ctingasnavigationdatumline.Inthisthesis,weexploreextractionalgorithmandoptimizationofcurvenavigationline.Themaincontentsareasfollows:1.Imagecapturingandprocessing.Imagesegmentationisthecrucialpartinimageprocessing.Excess·greenfeatureandthresholdsegmentationa
7、reusedtogetthebinaryimage.Thenwiththecombinationalgorithmofmathematicsmorphology,theeffectofsegmentationisimproved.2.Theextractionoffeaturepoints.Inordertoextractthenavigatingline,thenavigatingcharacteristicpointsarerequiredtobeascloseaspossibletothebeeline.
8、Fortheindefinitecurvatureofthecurvenavigatingline,newstandardofextractingisneeded.Thisthesisadoptsdirectionsofthemarchingcurveascharacteristicpoints.3.Thefittingofcurvenavigationline.Fitnavigati