[{"data":1,"prerenderedAt":318},["ShallowReactive",2],{"method-borrmann2008_6dlum":3},{"method":4,"reference":63,"equipment":86,"figures":115,"results":116},{"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":23,"limitations":29,"sensors":36,"platform":40,"estimator":43,"association":44,"timeModel":45,"deskew":46,"loopClosure":47,"globalOptimization":48,"mapRepresentation":49,"prior":50,"outputGeometry":51,"compute":52,"codeUrl":53,"codeLicense":54,"relatedVersions":55},"borrmann2008_6dlum","Borrmann et al., 2008","6D LUM (6-DoF Lu-Milios GraphSLAM)","Globally consistent 3D mapping with scan matching",2008,"classic","C04","full_slam_with_global_correction","本文把 Lu 與 Milios 的二維全域一致掃描對齊（每幅掃描一個位姿、以相對位姿關係構成網路、以最大概似同時求解）推廣到三維點雲與六自由度位姿，作者稱為 LUM，是直接建立在原始掃描對應點上的 GraphSLAM。每條連結的相對位姿與其共變異數由 ICP 找到的點對經泰勒展開線性化求得，再組成線性方程組以 Cholesky 分解求解，並反覆迭代、每次重算對應點與圖的連結。系統流程是先以 ICP 逐幅配準，當兩個估計位姿距離小於門檻時加入迴圈連結，再啟動 LUM 做全域鬆弛。作者另在網站公開勘誤附錄，修正線性化矩陣在左手座標系下的形式。","Extends Lu-Milios maximum-likelihood global scan alignment to 6-DoF poses and 3D scans (LUM), deriving linearized pose relations and covariances from ICP point pairs and solving the pose graph by Cholesky decomposition inside an ICP plus distance-based loop-closing pipeline.","full_text_reviewed","peer_reviewed_published","background","未在營建工地測試。Horn 資料集為 RIEGL 靜態地面雷射掃描的城鎮廣場（含教堂與紀念碑），參考值來自已知座標反光標靶的人工對齊（Sec. 7.1），性質上是既有建成環境的多站點雲全域一致化；與工程上多站 TLS、多段 SLAM 點雲整合時的全域平差問題相近（推論）。",[20,21,22],"completed_building","independent_reference","cross_site",[24,25,26,27,28],"Horn data set (13 TLS scans, artificially perturbed initial errors of 0.451-2.151 m and 1-5 degrees): position errors after registration 0.017-0.082 m and rotation errors below 0.1 degree against target-based reference alignment (Table 2, Sec. 7.1)","Pose errors are reduced evenly across scans without accumulation (Sec. 7.1)","Dynamic recomputation of the pose network lets the robot path merge correctly at the fourth loop in the 468-scan Hannover run (Sec. 7.2, Fig. 6-7)","Instabilities reported by earlier work for 2D scans with 3D poses did not appear in the 3D scan \u002F 6D pose case (Sec. 8)","Provides pose covariances in all six DoF (abstract, Sec. 6)",[30,31,32,33,34,35],"Loop closing relies on a simple distance heuristic; if it fails the map is incorrect, and a 6-DoF multi-hypothesis alternative was judged intractable (Sec. 5)","No ground truth for the 750 m Hannover run, so its final map could not be compared to a reference model; the authors cite Wulf et al. [40], who benchmarked the algorithm on a similar (not the same) data set against 2D land-registry maps and found higher precision where loops are closed (Sec. 7.2)","Matrix solution scales as O(n^3) in the number of scans (Sec. 6.5)","Assumes the robot stops for each scan (Sec. 6.1)","Linearization matrices printed in the article were later corrected in an author-posted addendum (addendum PDF)","Follow-up work: BALM states that such pairwise-registration-plus-pose-graph methods need repeated pairwise registration among all overlapping scans and do not optimize the point-cloud map directly, limiting attainable mapping consistency (balm2021 Sec. II, refs. [18]-[24])",[37,38,39],"robot-mounted 3D laser range finder, model not stated anywhere in the article (Hannover data set: 468 scans of 14,000 to 18,000 points, provided by Leibniz Universität Hannover)","high-resolution 3D laser scans provided by RIEGL LMS GmbH (Horn data set, 13 scans of 240,000 to 300,000 points); scanner model not stated; stationary acquisition is inferred from the stop-and-scan formulation and the target-based reference","planar 3-DoF odometry (x, z, theta_y) extrapolated to 6-DoF using the registration matrices of previously registered scans; the odometry source (for example wheel encoders) is not specified",[41,42],"mobile robot, locomotion type and model not stated (Hannover robot run 'driving' about 750 m)","static TLS stations (Horn main square, Sec. 7.1)","maximum-likelihood pose-graph estimation in the style of Lu and Milios ('LUM'): each link's pose relation and covariance are obtained by linearizing (Taylor expansion) the pose-compounding error of ICP point pairs, the linear system GX = B is solved by Cholesky decomposition, and the process is iterated with correspondences recomputed (Sec. 6.1-6.5)","ICP closest-point pairs within a distance limit (Sec. 4); graph links between scans whose estimated poses are closer than a threshold (5 m, Sec. 5; 7.5 m in the Hannover experiment, Sec. 7.2)","discrete poses; the robot stops to take each scan (Sec. 6.1)","avoided by design (stop-and-scan assumption, Sec. 6.1)","purely distance-based link creation between estimated poses; no appearance check; authors state that a failed loop closure yields an incorrect map (Sec. 5)","6-DoF LUM over all scan poses, O(n^3) per iteration for the 6n x 6n system; network links recomputed after each iteration so converging scans gain links and diverging ones lose them (Sec. 6.5-6.6, Sec. 7.2)","registered 3D point clouds","none (odometry only as initial guess)","globally consistent registered 3D point cloud with 6-DoF poses and their covariances (abstract, Sec. 6)","Horn: 13 scans of 240,000 to 300,000 points converged in 19 min with reduced points (6 min ICP, 13 min LUM; convergence when no scan moves more than 0.5 cm per iteration) (Sec. 7.1); closest-point search (step 2 of Algorithm 1, k-d tree, O(N log N) per link) dominates run time for n \u003C 13, N \u003C 300,000 and n \u003C 468, N \u003C 18,000 (Sec. 6.6); Cholesky solution O(n^3) (Sec. 6.5); hardware not reported anywhere in the article","https:\u002F\u002Fslam6d.sourceforge.io\u002F","GPL-3.0 (3DTK LICENSING file and package.xml in github.com\u002FJMUWRobotics\u002F3DTK, main branch; bundled third-party libraries keep their own licences; the file also places 3DTK-generated images and videos under CC BY-SA 3.0 and asks that the software not be used in military contexts). Applies to the current toolkit, not to a frozen release matching the paper.",[56,60],{"relation":57,"title":58,"doi_or_url":59},"erratum","Author-posted addendum (not a publisher erratum; no Crossref update record found) giving corrected linearization matrices H and M_i for the left-handed coordinate system","https:\u002F\u002Frobotik.informatik.uni-wuerzburg.de\u002Ftelematics\u002Fdownload\u002Fras2007_addendum.pdf",{"relation":61,"title":62,"doi_or_url":53},"code_release","3DTK - The 3D Toolkit (slam6d); its publications page lists this paper among those the toolkit implements",{"id":5,"kind":64,"shortName":7,"title":8,"authors":65,"year":9,"venue":71,"venueType":72,"publisher":73,"volumeIssuePages":74,"doi":75,"arxivId":76,"url":77,"firstPublicDate":78,"publicationStatus":16,"metadataStatus":79,"fulltextStatus":15,"era":10,"classicReason":80,"codeUrl":53,"cluster":11,"topics":81,"mdpi":82,"verification":83,"label":6,"fulltextRoute":84,"versionRead":85,"addedByCensus":82},"method",[66,67,68,69,70],"Dorit Borrmann","Jan Elseberg","Kai Lingemann","Andreas Nüchter","Joachim Hertzberg","Robotics and Autonomous Systems","journal","Elsevier","56(2): 130-142","10.1016\u002Fj.robot.2007.07.002",null,"http:\u002F\u002Fkos.informatik.uni-osnabrueck.de\u002Fdownload\u002Fras2007.pdf","2007-07-18","metadata_verified","principle reused: extends the Lu-Milios maximum-likelihood pose-network relaxation (lu_milios1997) to 6-DoF poses and full 3D scans, i.e. GraphSLAM directly on raw scan correspondences; BALM (balm2021 Sec. II, refs. [18]-[24] incl. [22]) cites it among pairwise-registration-plus-pose-graph methods that lidar bundle adjustment seeks to improve, which makes it the classic reference point for global consistency of 3D point-cloud maps.",[11],false,"corrected","author copy","version of record layout (Elsevier typeset PDF, Robotics and Autonomous Systems 56 (2008) 130-142) hosted on the authors' institutional server; author-posted addendum was checked in the earlier verification pass, not re-read here",[87,94,100,105,110],{"category":88,"model":89,"canonical":89,"role":90,"dataset":91,"specs":92,"locator":93},"lidar","3D laser scanner (model not stated; scans provided courtesy of RIEGL LMS GmbH)","dataset sensor","Horn (Austria) main square","13 high-resolution 3D scans of 240,000 to 300,000 points each; instrument type and model not stated (static terrestrial scanning is an inference from the stop-and-scan formulation and target-based reference)","Sec. 7.1; Fig. 2-3 captions; ref. [26]",{"category":95,"model":96,"canonical":96,"role":97,"dataset":91,"specs":98,"locator":99},"other","retro-reflective targets of known coordinates","reference or ground truth","used for manual alignment that serves as ground truth","Sec. 7.1",{"category":88,"model":101,"canonical":101,"role":90,"dataset":102,"specs":103,"locator":104},"robot-mounted 3D laser range finder (model not stated)","Hannover (Leibniz Universität Hannover robot run, about 750 m)","468 scans of 14,000 to 18,000 points each; stop-and-scan acquisition","Sec. 6.1; Sec. 7.2; Acknowledgments",{"category":95,"model":106,"canonical":106,"role":107,"dataset":76,"specs":108,"locator":109},"robot odometry (x, z, theta_y), source not stated","method input","planar 3-DoF odometry used only to extrapolate the initial 6-DoF pose","Sec. 3",{"category":111,"model":112,"canonical":112,"role":90,"dataset":102,"specs":113,"locator":114},"platform","mobile robot (Leibniz Universität Hannover; type not stated)","robot run of about 750 m","Sec. 7.2",[],{"totalRows":117,"groupCount":118,"groups":119,"others":317},55,2,[120,286],{"slug":121,"group":122,"sourceId":5,"sourceLabel":6,"table":123,"selfRows":124,"metrics":125,"seqs":137,"entrants":166,"cells":170,"outcomes":279,"locators":281,"hardware":282,"wordings":283,"notes":284},"borrmann2008-6dlum-table-2","borrmann2008_6dlum:Table 2","Table 2",52,[126,130,133,135],{"label":127,"unit":128,"statistic":129,"alignment":129},"Position error (m) after registration (Eucl. distance)","m","not_reported",{"label":131,"unit":132,"statistic":129,"alignment":129},"Rotation error (deg) after registration, theta_x component","deg",{"label":134,"unit":132,"statistic":129,"alignment":129},"Rotation error (deg) after registration, theta_y component",{"label":136,"unit":132,"statistic":129,"alignment":129},"Rotation error (deg) after registration, theta_z component",[138,142,144,146,148,150,152,154,156,158,160,162,164],{"dataset":139,"sequence":140,"environment":141},"Horn (Austria) main square, RIEGL scans","scan 1 (initial error 0.000 m; initial rotation (0.0, 0.0, 0.0) deg)","outdoor town square with church and monument (existing built environment)",{"dataset":139,"sequence":143,"environment":141},"scan 2 (initial error 0.780 m; initial rotation (0.0, 4.0, 0.0) deg)",{"dataset":139,"sequence":145,"environment":141},"scan 3 (initial error 1.965 m; initial rotation (0.0, 4.0, 0.0) deg)",{"dataset":139,"sequence":147,"environment":141},"scan 4 (initial error 1.764 m; initial rotation (0.0, 2.0, 0.0) deg)",{"dataset":139,"sequence":149,"environment":141},"scan 5 (initial error 1.222 m; initial rotation (0.0, 5.0, 0.0) deg)",{"dataset":139,"sequence":151,"environment":141},"scan 6 (initial error 0.859 m; initial rotation (0.0, 1.0, 0.0) deg)",{"dataset":139,"sequence":153,"environment":141},"scan 7 (initial error 2.151 m; initial rotation (0.0, 2.0, 0.0) deg)",{"dataset":139,"sequence":155,"environment":141},"scan 8 (initial error 0.451 m; initial rotation (0.0, 1.0, 0.0) deg)",{"dataset":139,"sequence":157,"environment":141},"scan 9 (initial error 1.023 m; initial rotation (0.0, 3.0, 0.0) deg)",{"dataset":139,"sequence":159,"environment":141},"scan 10 (initial error 2.039 m; initial rotation (0.0, 1.0, 0.0) deg)",{"dataset":139,"sequence":161,"environment":141},"scan 11 (initial error 1.654 m; initial rotation (0.0, 5.0, 0.0) deg)",{"dataset":139,"sequence":163,"environment":141},"scan 12 (initial error 1.340 m; initial rotation (0.0, 2.0, 0.0) deg)",{"dataset":139,"sequence":165,"environment":141},"scan 13 (initial error 1.195 m; initial rotation (0.0, 5.0, 0.0) deg)",[167],{"name":168,"methodId":5,"linkable":169,"proposed":169,"self":169},"6D LUM (ICP + Lu-Milios style GraphSLAM)",true,[171,174,176,177,179,181,183,185,187,189,191,193,195,197,198,200,202,205,207,209,211,214,216,218,220,223,225,227,229,232,234,236,238,241,243,245,246,249,250,252,254,256,258,260,262,265,267,269,271,273,275,277],[172,172,172,172,172,172,173,173,172],0,-1,[172,175,172,172,172,172,173,173,172],1,[172,118,172,172,172,172,173,173,172],[172,178,172,172,172,172,173,173,172],3,[172,172,175,180,173,172,173,173,172],0.017,[172,175,175,182,173,172,173,173,172],-0.0221,[172,118,175,184,173,172,173,173,172],0.0177,[172,178,175,186,173,172,173,173,172],-0.0116,[172,172,118,188,173,172,173,173,172],0.024,[172,175,118,190,173,172,173,173,172],-0.0242,[172,118,118,192,173,172,173,173,172],0.0257,[172,178,118,194,173,172,173,173,172],-0.0236,[172,172,178,196,173,172,173,173,172],0.029,[172,175,178,190,173,172,173,173,172],[172,118,178,199,173,172,173,173,172],0.0228,[172,178,178,201,173,172,173,173,172],-0.02,[172,172,203,204,173,172,173,173,172],4,0.037,[172,175,203,206,173,172,173,173,172],-0.0266,[172,118,203,208,173,172,173,173,172],0.0298,[172,178,203,210,173,172,173,173,172],-0.0339,[172,172,212,213,173,172,173,173,172],5,0.059,[172,175,212,215,173,172,173,173,172],-0.0297,[172,118,212,217,173,172,173,173,172],0.0183,[172,178,212,219,173,172,173,173,172],0.0039,[172,172,221,222,173,172,173,173,172],6,0.052,[172,175,221,224,173,172,173,173,172],0.0061,[172,118,221,226,173,172,173,173,172],0.0195,[172,178,221,228,173,172,173,173,172],-0.0269,[172,172,230,231,173,172,173,173,172],7,0.053,[172,175,230,233,173,172,173,173,172],-0.0416,[172,118,230,235,173,172,173,173,172],0.0173,[172,178,230,237,173,172,173,173,172],0.0076,[172,172,239,240,173,172,173,173,172],8,0.069,[172,175,239,242,173,172,173,173,172],-0.069,[172,118,239,244,173,172,173,173,172],0.0279,[172,178,239,180,173,172,173,173,172],[172,172,247,248,173,172,173,173,172],9,0.082,[172,175,247,242,173,172,173,173,172],[172,118,247,251,173,172,173,173,172],0.0165,[172,178,247,253,173,172,173,173,172],0.019,[172,172,255,213,173,172,173,173,172],10,[172,175,255,257,173,172,173,173,172],-0.0988,[172,118,255,259,173,172,173,173,172],0.0936,[172,178,255,261,173,172,173,173,172],0.004,[172,172,263,264,173,172,173,173,172],11,0.033,[172,175,263,266,173,172,173,173,172],0.003,[172,118,263,268,173,172,173,173,172],0.0196,[172,178,263,270,173,172,173,173,172],0.041,[172,172,272,180,173,172,173,173,172],12,[172,175,272,274,173,172,173,173,172],0.0022,[172,118,272,276,173,172,173,173,172],0.0593,[172,178,272,278,173,172,173,173,172],0.051,[280],"reference scan (no perturbation)",[123],[],[],[285],"Horn TLS data set, 13 scans; poses perturbed with random initial errors (horizontal position, vertical-axis rotation), then ICP + 6D LUM; errors after registration against ground truth from manual alignment with retro-reflective targets of known coordinates; scan 1 is the unperturbed reference",{"slug":287,"group":288,"sourceId":5,"sourceLabel":6,"table":289,"selfRows":178,"metrics":290,"seqs":299,"entrants":303,"cells":305,"outcomes":311,"locators":312,"hardware":313,"wordings":314,"notes":315},"borrmann2008-6dlum-text-sec-7-1","borrmann2008_6dlum:Text Sec. 7.1","Text Sec. 7.1",[291,295,297],{"label":292,"unit":293,"statistic":129,"alignment":294},"total scan-matching time until convergence (ICP + LUM)","min","not_applicable",{"label":296,"unit":293,"statistic":129,"alignment":294},"ICP time",{"label":298,"unit":293,"statistic":129,"alignment":294},"LUM time",[300],{"dataset":139,"sequence":301,"environment":302},"all 13 scans","outdoor town square",[304],{"name":168,"methodId":5,"linkable":169,"proposed":169,"self":169},[306,308,309],[172,172,172,307,173,172,173,173,172],19,[172,175,172,221,173,172,173,173,172],[172,118,172,310,173,172,173,173,172],13,[],[99],[],[],[316],"Horn data set, 13 scans of 240,000 to 300,000 points, matching with reduced points until no scan moved more than 0.5 cm per iteration",[],1790510655882]