資源描述:
《用于公路勘察設(shè)計(jì)的機(jī)載lidar點(diǎn)云抽稀研究》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫。
1、分類號(hào):P2310710-2015126045碩士學(xué)位論文用于公路勘察設(shè)計(jì)的機(jī)載LiDAR點(diǎn)云抽稀研究方程喜導(dǎo)師姓名職稱隋立春教授申請(qǐng)學(xué)位類別工學(xué)碩士學(xué)科專業(yè)名稱攝影測(cè)量與遙感論文提交日期2018年5月1日論文答辯日期2018年6月3日學(xué)位授予單位長(zhǎng)安大學(xué)ResearchonAirborneLiDARPointCloudThinningforHighwaySurveyandDesignAThesisSubmittedfortheDegreeofMasterCandidate:FangChengxiSup
2、ervisor:Prof.SuiLichunChang’anUniversity,Xi’an,China摘要在公路建設(shè)領(lǐng)域,從公路初建的地形測(cè)繪到公路改擴(kuò)建,機(jī)載LiDAR技術(shù)都能提供高精度的地形數(shù)據(jù),然而,由于公路測(cè)區(qū)狹長(zhǎng)、LiDAR點(diǎn)云密度高等因素,使得機(jī)載LiDAR獲取的數(shù)據(jù)量非常龐大,龐大的數(shù)據(jù)量會(huì)影響公路勘察設(shè)計(jì)中DEM的構(gòu)建速度、數(shù)據(jù)交互的難易程度和數(shù)據(jù)的存儲(chǔ)速度,所以對(duì)用于公路勘察設(shè)計(jì)的機(jī)載LiDAR點(diǎn)云進(jìn)行抽稀精簡(jiǎn)有利于實(shí)際工程應(yīng)用。關(guān)于LiDAR點(diǎn)云的抽稀方法,如何在保留地形特征點(diǎn)的同
3、時(shí),保證點(diǎn)位分布的合理性,避免出現(xiàn)大面積的點(diǎn)云空洞,以及提高算法的處理速度,仍然是當(dāng)下研究的重點(diǎn)與難點(diǎn)。本文主要研究?jī)?nèi)容如下:(1)總結(jié)、分析了規(guī)則格網(wǎng)、八叉樹和KD樹三種海量點(diǎn)云數(shù)據(jù)空間索引方法的優(yōu)缺點(diǎn)。結(jié)合機(jī)載LiDAR點(diǎn)云數(shù)據(jù)的特點(diǎn),采用KD樹作為本文實(shí)驗(yàn)數(shù)據(jù)的空間索引結(jié)構(gòu),提高算法處理速度。(2)根據(jù)機(jī)載LiDAR的工作原理,分析總結(jié)其主要的粗差來源,研究采用基于假設(shè)檢驗(yàn)的方法剔除粗差。(3)提出了基于平均曲率的點(diǎn)云抽稀算法,并利用標(biāo)記法來解決點(diǎn)云空洞的問題,提高點(diǎn)位分布的合理性,盡可能的保證抽
4、稀后的點(diǎn)云精度。(4)設(shè)計(jì)開發(fā)了點(diǎn)云顯示、點(diǎn)云粗差剔除、點(diǎn)云抽稀等功能的軟件,并基于實(shí)際數(shù)據(jù)進(jìn)行點(diǎn)云粗差剔除和抽稀實(shí)驗(yàn),對(duì)實(shí)驗(yàn)結(jié)果進(jìn)行對(duì)比分析。關(guān)鍵詞:機(jī)載LiDAR,空間索引,粗差剔除,抽稀算法,曲率iAbstractInthefieldofhighwayconstruction,fromtheinitialmappingofroadstothereconstructionandexpansionofhighways,airborneLiDARtechnologycanprovidehigh-prec
5、isionterraindata,however,duetothenarrownessofthehighwaysurveyareaandthehighdensityofLiDARpointcloud,theamountofdataacquiredbytheairborneLiDARishuge,thehugeamountofdatawillaffectthespeedofDEMconstruction,easeofdataexchange,anddatastoragespeedinhighwaysurv
6、eyanddesign,therefore,itisusefulforpracticalengineeringapplicationtothinningandsimplifyairborneLiDARpointcloudusedforhighwaysurveyanddesign.WithrespecttothemethodofthinningLiDARpointclouds,howtoensuretherationalityofthedistributionofpointswhilepreserving
7、theterrainfeaturepoints,avoidtheoccurrenceoflarge-areapointcloudholes,andimprovetheprocessingspeedofthealgorithmisstillthefocusanddifficultyofcurrentresearch.Inthispaper,themainresearchcontentsareasfollows:(1)Theadvantagesanddisadvantagesofthethreekindso
8、fmassivepointclouddataspatialindexingmethodsofregulargrid,octree,andKDtreearesummarizedandanalyzed.CombiningthecharacteristicsofairborneLiDARpointclouddata,theKDtreeisusedasthespatialindexstructureoftheexperimentaldatainth