[{"data":1,"prerenderedAt":142},["ShallowReactive",2],{"method-gutmann_konolige1999_lrgc":3},{"method":4,"reference":52,"equipment":72,"figures":86,"results":87},{"id":5,"label":6,"shortName":7,"title":8,"year":9,"era":10,"cluster":11,"scope":12,"keyIdeaZh":13,"keyIdeaEn":14,"fulltextStatus":15,"publicationStatus":16,"recommendation":17,"constructionRelevance":18,"validationEnvironment":19,"strengths":21,"limitations":25,"sensors":30,"platform":33,"estimator":35,"association":36,"timeModel":37,"deskew":38,"loopClosure":39,"globalOptimization":40,"mapRepresentation":41,"prior":42,"outputGeometry":43,"compute":44,"codeUrl":45,"codeLicense":46,"relatedVersions":47},"gutmann_konolige1999_lrgc","Gutmann & Konolige, 1999","LRGC (Local Registration and Global Correlation)","Incremental mapping of large cyclic environments",1999,"classic","C01","full_slam_with_global_correction","LRGC 以 Lu 與 Milios 的一致位姿估計為核心，分兩種方式使用：每加入一筆新掃描，只與最近 K 個位姿做局部配準，所以每步計算量固定；偵測到迴圈後，才對整個迴圈做一致位姿估計。迴圈偵測不用單一掃描，而是把最新 m 筆掃描組成地圖片段，以相關運算在較舊的地圖中搜尋，並以高匹配分數、低歧異與低變異三個條件拒絕誤判；搜尋範圍與片段大小都隨位置不確定度增加。作者稱這是第一個不需操作者輸入即可在大型循環環境即時產生稠密度量地圖的系統。","Incremental 2D laser mapping with constant-time local Lu-Milios registration over the last K poses and correlation of multi-scan map patches for loop detection, followed by consistent pose estimation over the closed loop.","full_text_reviewed","peer_reviewed_published","background","未在營建場域驗證；資料來自 CMU Wean Hall、SRI 人工智慧中心、卡內基自然史博物館與 Freiburg 人工智慧實驗室等既有建築。作者以多筆掃描組成的地圖片段做迴圈偵測，理由是單筆掃描在沿走廊方向變化少時難以排除誤判（Sec. 1.4），這與大型建築長走廊與重複樓層的建圖風險直接相關（推論）。",[20],"completed_building",[22,23,24],"For K >= 7, local registration differed from full registration by less than 1 mm average error per pose on a 150-pose map (Sec. 2.2; Fig. 3).","Closed a cycle of about 200 m in Wean Hall (80 m x 25 m, two cycles) and corrected strong directional-carpet odometry drift in an 85 m x 15 m SRI environment (Sec. 3; Figs. 7-8).","Constant-time incremental updates; large loops closed in less than 10 s in almost all cases (Sec. 2.2).",[26,27,28,29],"A topological connection, once made, cannot be undone, so false positives are critical (Sec. 2.3).","Requires good scan-matching results; the authors report that stereo-camera range data was much less accurate, but the rest of that passage is missing from the VoR scan (Sec. 4).","If the filters reject too many good matches loops cannot be closed, and if they accept a false match the map becomes inconsistent; multiple loop hypotheses are left for future work (Sec. 4, author copy).","Local registration can sometimes differ radically from global registration, which the authors did not examine closely (Sec. 2.2).",[31,32],"2D laser range finder (180 deg SICK)","wheel odometry",[34],"wheeled UGV (B21, Pioneer II, Pioneer I)","Lu-Milios consistent pose estimation (least-squares pose network) applied to the last K poses for each new scan and, after a loop is found, to the poses along the loop with sparse linear algebra and strong-link reduction (maximum 200 poses) (Sec. 2.2)","pairwise scan matching combining Cox's point-to-line method with Lu-Milios matching (Gutmann-Schlegel method); loop detection by correlating a patch of the newest m scans with the older map on a grid, accepted only if match score is high, ambiguity low and variance low (Sec. 2.1, 2.3)","discrete poses (new scan added after about 20 to 50 cm of motion)","not_reported","patch-to-map correlation run every few scans inside a Mahalanobis-gated search area; patch size grows with position uncertainty; an accepted topological link cannot be undone (Sec. 2.3-2.4)","consistent pose estimation over the closed loop, rerun with new scan matches between newly linked poses; optional final optimization over all poses at the end of a run (Sec. 2.2, 2.4)","undirected graph of robot poses with attached scans; links from dead reckoning, scan matching or correlation (Sec. 2.4)","none","2D dense metric scan-point maps","incremental registration under 100 ms in typical circumstances; closing even large loops under 10 s in almost all cases (Sec. 2.2); constant time except when closing loops",null,"not_applicable",[48],{"relation":49,"title":50,"doi_or_url":51},"author_copy","Author pre-print on Konolige's SRI page (conclusion wording differs slightly from the VoR)","http:\u002F\u002Fwww.ai.sri.com\u002F~konolige\u002Fpapers\u002Fmapping.pdf",{"id":5,"kind":53,"shortName":7,"title":8,"authors":54,"year":9,"venue":57,"venueType":58,"publisher":59,"volumeIssuePages":60,"doi":61,"arxivId":45,"url":62,"firstPublicDate":63,"publicationStatus":16,"metadataStatus":64,"fulltextStatus":15,"era":10,"classicReason":65,"codeUrl":45,"cluster":11,"topics":66,"mdpi":67,"verification":68,"label":6,"fulltextRoute":69,"versionRead":70,"addedByCensus":71},"method",[55,56],"Jens-Steffen Gutmann","Kurt Konolige","Proceedings 1999 IEEE International Symposium on Computational Intelligence in Robotics and Automation (CIRA'99), Monterey, CA","conference","IEEE","pp. 318-325","10.1109\u002Fcira.1999.810068","https:\u002F\u002Fdoi.org\u002F10.1109\u002FCIRA.1999.810068","1999-11","metadata_verified","principle reused: early incremental graph-style laser mapping that keeps local Lu-Milios registration constant-time and closes loops by correlating multi-scan map patches with false-positive filters; Hähnel et al. [hahnel2003_gridfastslam] and the SPA paper [karto_spa2010] cite it for loop closing and graph construction.",[11],false,"corrected","NTU institutional (curl)","IEEE Xplore version of record (scanned PDF with OCR text layer, 8 pp.); author pre-print from www.ai.sri.com\u002F~konolige\u002Fpapers\u002Fmapping.pdf read for the conclusion passage missing from the VoR scan",true,[73,78,82,84],{"category":74,"model":75,"canonical":75,"role":76,"dataset":45,"specs":38,"locator":77},"platform","B21","method input","Sec. 3",{"category":79,"model":80,"canonical":80,"role":76,"dataset":45,"specs":81,"locator":77},"lidar","SICK laser range finder","180 deg field of view",{"category":74,"model":83,"canonical":83,"role":76,"dataset":45,"specs":38,"locator":77},"Pioneer II",{"category":74,"model":85,"canonical":85,"role":76,"dataset":45,"specs":38,"locator":77},"Pioneer I",[],{"totalRows":88,"groupCount":89,"groups":90,"others":141},3,1,[91],{"slug":92,"group":93,"sourceId":5,"sourceLabel":6,"table":94,"selfRows":88,"metrics":95,"seqs":106,"entrants":112,"cells":119,"outcomes":128,"locators":132,"hardware":135,"wordings":136,"notes":137},"gutmann-konolige1999-lrgc-text-sec-2-2","gutmann_konolige1999_lrgc:Text Sec. 2.2","Text Sec. 2.2",[96,100,103],{"label":97,"unit":98,"statistic":99,"alignment":42},"average error per pose for K >= 7 (smaller than a millimeter)","mm","mean",{"label":101,"unit":102,"statistic":38,"alignment":42},"incremental registration time (under 100 ms)","ms",{"label":104,"unit":105,"statistic":38,"alignment":42},"computation for closing even large loops (less than 10 seconds in almost all cases)","s",[107,111],{"dataset":108,"sequence":109,"environment":110},"150-pose map (source log not identified in the paper; about 0.3 m between poses)","150-pose map","indoor",{"dataset":38,"sequence":38,"environment":110},[113,115,117],{"name":114,"methodId":5,"linkable":71,"proposed":71,"self":71},"LRGC local registration (K >= 7)",{"name":116,"methodId":5,"linkable":71,"proposed":71,"self":71},"LRGC local registration",{"name":118,"methodId":5,"linkable":71,"proposed":71,"self":71},"LRGC loop closing",[120,123,125],[121,121,121,89,121,121,122,122,121],0,-1,[89,89,89,124,89,89,122,122,89],100,[126,126,89,127,126,89,122,122,126],2,10,[129,130,131],"upper bound: below 1 mm","upper bound: under 100 ms in typical circumstances","upper bound in almost all cases",[133,134],"Sec. 2.2; Fig. 3; footnote 1","Sec. 2.2",[],[],[138,139,140],"Local registration over the last K poses versus update of all poses on a 150-pose map (about 0.3 m between poses); pose error from incremental differences between consecutive poses","Incremental registration time in typical circumstances","Loop closing with sparse solvers and strong-link reduction (maximum 200 poses)",[],1790510661738]