[{"data":1,"prerenderedAt":371},["ShallowReactive",2],{"method-elasticlidarfusion2018":3},{"method":4,"reference":57,"equipment":80,"figures":113,"results":114},{"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":29,"platform":34,"estimator":36,"association":37,"timeModel":38,"deskew":39,"loopClosure":40,"globalOptimization":41,"mapRepresentation":42,"prior":43,"outputGeometry":44,"compute":45,"codeUrl":46,"codeLicense":47,"relatedVersions":48},"elasticlidarfusion2018","Park et al., 2018","Elastic LiDAR Fusion","Elastic LiDAR Fusion: Dense Map-Centric Continuous-Time SLAM",2018,"recent","C04","full_slam_with_global_correction","Elastic LiDAR Fusion 把連續時間（continuous-time）SLAM 與 ElasticFusion 的「以地圖為中心」（map-centric）概念結合：局部仍以滑動視窗的連續時間軌跡處理手持旋轉 LiDAR 的運動畸變，但全域一致性不靠整條軌跡的批次最佳化，而是在迴圈發生時對整張面元地圖做非剛性變形（deformation graph）。因此迴圈閉合的計算量取決於迴圈前探索的空間大小，而非運作時間。多次觀測以機率式面元融合（surfel fusion）合併，論文報告可降低重建表面的雜訊。","Elastic LiDAR Fusion keeps a local continuous-time trajectory for de-skewing but achieves global consistency by deforming a probabilistically fused surfel map on loop closure instead of batch trajectory optimization.","full_text_reviewed","peer_reviewed_published","main_body","not_reported（測試於辦公室、會議室、多樓層結構與室內外混合環境等既有建物；報告的平面補丁雜訊指標與工程表面品質相關（推論），但未做工地或工程任務驗證）。",[20],"completed_building",[22,23,24],"Loop-closure optimisation state dimension and time much smaller than batch CT-SLAM: 192 states and 0.12 s versus 3396 states and 195.40 s on the Fig. 1 office map (Sec. VII-A, Table I; hardware not reported)","Floor-patch projective-distance noise 5.79-13.07 mm vs 15.78-19.40 mm for the CT-SLAM point cloud (four 0.7 m patches) (Table III)","Uniform surfel density after fusion regardless of raw point count (Sec. VII-C)",[26,27,28],"Trajectory accuracy (RMSE 0.041-0.076 m) is measured against the batch CT-SLAM trajectory of ref. [3] (Bosse et al. 2012, Zebedee), not an independent ground truth (Sec. VII-B, Table II)","Surface evaluation uses planarity of floor patches (projective distance to each patch's mean plane) because ground truth is not available (Sec. VII-C)","Journal extension reports map distortion reaching 10 cm on a 60 m scale map and a partial-observation problem (elasticity_ct2022, Sec. IX, arXiv version)",[30,31,32,33],"rotating 2D LiDAR (Hokuyo UTM-30LX spinning, with encoder)","IMU (Microstrain 3DM-GX3)","Grasshopper3 2.8 MP colour camera with fisheye lens used only for colourisation","Optris PI 450 thermal-infrared camera (382 x 288 pixels) on the device but not used",[35],"handheld","local sliding-window continuous-time trajectory optimisation solved by iterative nonlinear least squares over subsampled trajectory elements Q, IMU biases and an additional state d that the paper does not define (x = [Q, b_omega, b_alpha, d]), combining surfel-to-surfel, surfel-to-map-prior and IMU acceleration and angular-velocity constraints (Eq. 3-8); global consistency by Gauss-Newton optimisation of an ElasticFusion-style deformation graph with loop, pinning and regularisation terms (Eq. 16-19)","multi-resolution 3D ellipsoidal sparse surfels (from Bosse and Zlot 2009, ref. [9]) matched pairwise and to the global-map prior in their averaged normal direction (Eq. 4-5); dense 2D disk surfels associated with a sensor-noise model that searches deeper along the beam direction under a resolution threshold, then fused by Bayesian fusion (Sec. V-A)","continuous-time trajectory in a local window using linear interpolation between subsampled poses (slerp-like rotation, linear translation), chosen over B-splines to keep high-frequency motion at low cost","handled by the continuous-time trajectory representation (Sec. I, VIII)","two detection sources: (1) rigid ICP between active and inactive sparse-surfel maps detects moderate misalignment on the fly (Algorithm 1); (2) for large misalignment, 3D point-cloud place recognition by keypoint voting (ref. [17], Bosse and Zlot 2013) with descriptors computed every frame and compared with stored scene keys; loop constraints are applied as map deformation, not trajectory optimisation","deformation graph applied to the whole map (ElasticFusion-inspired) with surfel uncertainty propagation (Sec. III)","sparse multi-resolution ellipsoid surfel map plus dense 2D disk surfel map with probabilistic surfel fusion (Sec. III)","no external prior; the initial map prior is built from a short period of stationary scanning at the start","dense fused surfel map (e.g., 20 mm surfels at 10 mm resolution in Fig. 1); no global trajectory is maintained (Sec. VII-B)","per-frame runtime and hardware not reported anywhere in the paper; only global loop-closure optimisation cost is given (192 states and 0.12 s for the proposed method versus 3396 states and 195.40 s for batch CT-SLAM on the Fig. 1 map, Table I)",null,"not_verified",[49,53],{"relation":50,"title":51,"doi_or_url":52},"preprint","arXiv 1711.01691 (v1 2017-11-06, v3 2018-03-05)","https:\u002F\u002Farxiv.org\u002Fabs\u002F1711.01691",{"relation":54,"title":55,"doi_or_url":56},"journal_extension","Elasticity Meets Continuous-Time: Map-Centric Dense 3D LiDAR SLAM","10.1109\u002FTRO.2021.3096650",{"id":5,"kind":58,"shortName":7,"title":8,"authors":59,"year":9,"venue":66,"venueType":67,"publisher":68,"volumeIssuePages":69,"doi":70,"arxivId":71,"url":52,"firstPublicDate":72,"publicationStatus":16,"metadataStatus":73,"fulltextStatus":15,"era":10,"classicReason":74,"codeUrl":46,"cluster":11,"topics":75,"mdpi":76,"verification":77,"label":6,"fulltextRoute":78,"versionRead":79,"addedByCensus":76},"method",[60,61,62,63,64,65],"Chanoh Park","Peyman Moghadam","Soohwan Kim","Alberto Elfes","Clinton Fookes","Sridha Sridharan","2018 IEEE International Conference on Robotics and Automation (ICRA)","conference","IEEE","pp. 1206-1213","10.1109\u002Ficra.2018.8462915","1711.01691","2017-11-06","metadata_verified","not_applicable",[11],false,"corrected","arXiv","arXiv 1711.01691v3 (2018-03-05; comments 'accepted to ICRA 2018'); IEEE version of record not compared because IEEE Xplore was temporarily unavailable during this session",[81,87,92,97,102,108],{"category":82,"model":83,"canonical":83,"role":84,"dataset":46,"specs":85,"locator":86},"lidar","Hokuyo UTM-30LX","method input","2D laser spun on a hand-held device to give 3D scans","Sec. VII; Fig. 2a",{"category":88,"model":89,"canonical":89,"role":84,"dataset":46,"specs":90,"locator":91},"other","encoder (spinning mechanism)","part of the hand-held spinning LiDAR","Sec. VII; Fig. 2a caption",{"category":93,"model":94,"canonical":94,"role":84,"dataset":46,"specs":95,"locator":96},"imu","Microstrain 3DM-GX3","not_reported","Sec. VII",{"category":98,"model":99,"canonical":99,"role":84,"dataset":46,"specs":100,"locator":101},"camera","Grasshopper3 2.8 MP color camera","2.8 MP; fisheye lens (Sec. V-A); used only for colourising the dense surfel map","Sec. V-A; Sec. VII",{"category":103,"model":104,"canonical":104,"role":105,"dataset":106,"specs":107,"locator":91},"thermal","Optris PI 450 thermal-infrared camera","dataset sensor","authors' hand-held spinning LiDAR data","382 x 288 pixels; mounted on the device; Fig. 2a caption states the thermal camera is not used in the paper",{"category":109,"model":110,"canonical":110,"role":84,"dataset":46,"specs":111,"locator":112},"mobile_scanner_device","experimental handheld 3D spinning LiDAR (as written; builder not stated)","integrates 2D laser, encoder, IMU, colour camera and thermal camera","Fig. 2a; Sec. VII",[],{"totalRows":115,"groupCount":116,"groups":117,"others":370},22,4,[118,251,288,336],{"slug":119,"group":120,"sourceId":121,"sourceLabel":122,"table":123,"selfRows":124,"metrics":125,"seqs":136,"entrants":145,"cells":156,"outcomes":244,"locators":245,"hardware":247,"wordings":248,"notes":249},"elasticity-ct2022-table-vi","elasticity_ct2022:Table VI","elasticity_ct2022","Park et al., 2022","Table VI",12,[126,130,132,135],{"label":127,"unit":128,"statistic":129,"alignment":74},"e_t translation error","m","RMSE",{"label":127,"unit":128,"statistic":131,"alignment":74},"std",{"label":133,"unit":134,"statistic":129,"alignment":74},"e_r rotation error (rotation vector norm)","rad",{"label":133,"unit":134,"statistic":131,"alignment":74},[137,141,143],{"dataset":138,"sequence":139,"environment":140},"authors' mixed indoor and outdoor point clouds","Easy initial guess","indoor and outdoor mixed",{"dataset":138,"sequence":142,"environment":140},"Medium initial guess",{"dataset":138,"sequence":144,"environment":140},"Hard initial guess",[146,149,152,154],{"name":147,"methodId":5,"linkable":148,"proposed":76,"self":148},"(a) Sparse surfel ICP (configuration of previous work [2])",true,{"name":150,"methodId":151,"linkable":148,"proposed":76,"self":76},"(b) Open3D global registration (FPFH + RANSAC) [60]","zhou2018open3d",{"name":153,"methodId":46,"linkable":76,"proposed":76,"self":76},"(c) SHOT initialisation + point-to-plane ICP [61]",{"name":155,"methodId":121,"linkable":148,"proposed":148,"self":76},"(d) Proposed sequential metric localisation",[157,161,163,166,168,170,172,174,176,178,180,182,184,185,186,188,190,192,194,196,198,200,202,203,205,207,209,211,213,214,216,218,219,221,223,225,226,228,230,231,233,235,237,238,240,241,242,243],[158,158,158,159,160,158,160,160,158],0,0.04,-1,[158,162,158,159,160,158,160,160,158],1,[158,164,158,165,160,158,160,160,158],2,0.01,[158,167,158,165,160,158,160,160,158],3,[162,158,158,169,160,158,160,160,158],0.3,[162,162,158,171,160,158,160,160,158],0.65,[162,164,158,173,160,158,160,160,158],0.03,[162,167,158,175,160,158,160,160,158],0.06,[164,158,158,177,160,158,160,160,158],1.49,[164,162,158,179,160,158,160,160,158],1.52,[164,164,158,181,160,158,160,160,158],0.07,[164,167,158,183,160,158,160,160,158],0.12,[167,158,158,173,160,158,160,160,158],[167,162,158,165,160,158,160,160,158],[167,164,158,187,160,158,160,160,158],0.001,[167,167,158,189,160,158,160,160,158],0.0005,[158,158,162,191,160,158,160,160,158],0.4,[158,162,162,193,160,158,160,160,158],0.51,[158,164,162,195,160,158,160,160,158],0.19,[158,167,162,197,160,158,160,160,158],0.31,[162,158,162,199,160,158,160,160,158],1.64,[162,162,162,201,160,158,160,160,158],2.39,[162,164,162,169,160,158,160,160,158],[162,167,162,204,160,158,160,160,158],0.38,[164,158,162,206,160,158,160,160,158],1.53,[164,162,162,208,160,158,160,160,158],1.55,[164,164,162,210,160,158,160,160,158],0.11,[164,167,162,212,160,158,160,160,158],0.29,[167,158,162,159,160,158,160,160,158],[167,162,162,215,160,158,160,160,158],0.02,[167,164,162,217,160,158,160,160,158],0.004,[167,167,162,187,160,158,160,160,158],[158,158,164,220,160,158,160,160,158],2.52,[158,162,164,222,160,158,160,160,158],0.87,[158,164,164,224,160,158,160,160,158],2.42,[158,167,164,208,160,158,160,160,158],[162,158,164,227,160,158,160,160,158],13.8,[162,162,164,229,160,158,160,160,158],22.4,[162,164,164,204,160,158,160,160,158],[162,167,164,232,160,158,160,160,158],0.63,[164,158,164,234,160,158,160,160,158],7.59,[164,162,164,236,160,158,160,160,158],18.56,[164,164,164,169,160,158,160,160,158],[164,167,164,239,160,158,160,160,158],0.71,[167,158,164,175,160,158,160,160,158],[167,162,164,159,160,158,160,160,158],[167,164,164,217,160,158,160,160,158],[167,167,164,187,160,158,160,160,158],[],[246],"Table VI (VoR, p. 993; identical to arXiv v1 Table V)",[],[],[250],"loop-closure misalignment estimation on mixed indoor and outdoor data; ground truth from the globally optimised trajectory; 10 locations x 50 random initial guesses (500 triggers) per level; Easy sigma_theta_z 10 deg, sigma_theta_xy 1 deg, sigma_t 0.5 m; Medium 50, 5, 5; Hard 100, 20, 50; text calls the values RMSE, caption calls them error norms with std in parentheses. Values read from the VoR Table VI by the second checker",{"slug":252,"group":253,"sourceId":5,"sourceLabel":6,"table":254,"selfRows":116,"metrics":255,"seqs":258,"entrants":270,"cells":273,"outcomes":282,"locators":283,"hardware":284,"wordings":285,"notes":286},"elasticlidarfusion2018-table-ii","elasticlidarfusion2018:Table II","Table II",[256],{"label":257,"unit":128,"statistic":129,"alignment":95},"Traj Error (m), absolute trajectory RMSE vs CT-SLAM [3]",[259,262,265,268],{"dataset":106,"sequence":260,"environment":261},"Fig. 1 (length 330 m, 14.6 min, size 20x20 m)","office",{"dataset":106,"sequence":263,"environment":264},"Fig. 7 (i) (length 130 m, 6.1 min, size 10x6 m)","small meeting room",{"dataset":106,"sequence":266,"environment":267},"Fig. 7 (ii) (length 300 m, 11.4 min, size 55x20 m)","multiple floor structure",{"dataset":106,"sequence":269,"environment":140},"Fig. 7 (iii) (length 360 m, 9.1 min, size 60x25 m)",[271],{"name":272,"methodId":5,"linkable":148,"proposed":148,"self":148},"Proposed (Elastic LiDAR Fusion)",[274,276,278,280],[158,158,158,275,160,158,160,160,158],0.047,[158,158,162,277,160,158,160,160,158],0.041,[158,158,164,279,160,158,160,160,158],0.056,[158,158,167,281,160,158,160,160,158],0.076,[],[254],[],[],[287],"absolute trajectory RMSE between the deformed trajectory of the proposed method and the globally optimised CT-SLAM [3] trajectory (reference is another estimate, not ground truth); trajectories stored only for this comparison",{"slug":289,"group":290,"sourceId":5,"sourceLabel":6,"table":291,"selfRows":116,"metrics":292,"seqs":297,"entrants":307,"cells":313,"outcomes":330,"locators":331,"hardware":332,"wordings":333,"notes":334},"elasticlidarfusion2018-table-iii","elasticlidarfusion2018:Table III","Table III",[293],{"label":294,"unit":295,"statistic":296,"alignment":74},"projective distance error to patch mean plane","mm","mean",[298,301,303,305],{"dataset":106,"sequence":299,"environment":300},"patch a (47.8x10^4 points, 3.7x10^3 surfels)","indoor floor (map of Fig. 8b)",{"dataset":106,"sequence":302,"environment":300},"patch b (37.8x10^4 points, 4.1x10^3 surfels)",{"dataset":106,"sequence":304,"environment":300},"patch c (40.6x10^4 points, 3.8x10^3 surfels)",{"dataset":106,"sequence":306,"environment":300},"patch d (56.3x10^4 points, 3.8x10^3 surfels)",[308,311],{"name":309,"methodId":310,"linkable":148,"proposed":76,"self":76},"CT-SLAM [3] (raw point cloud)","zebedee2012",{"name":312,"methodId":5,"linkable":148,"proposed":148,"self":148},"Proposed (Elastic LiDAR Fusion, fused surfels)",[314,316,318,320,322,324,326,328],[158,158,158,315,160,158,160,160,158],16.08,[162,158,158,317,160,158,160,160,158],7.72,[158,158,162,319,160,158,160,160,158],15.78,[162,158,162,321,160,158,160,160,158],5.79,[158,158,164,323,160,158,160,160,158],16.43,[162,158,164,325,160,158,160,160,158],10.39,[158,158,167,327,160,158,160,160,158],19.4,[162,158,167,329,160,158,160,160,158],13.07,[],[291],[],[],[335],"floor patches of 0.7 m radius (Fig. 8b); error = mean projective distance of points or surfels to the mean plane of each patch (relative noise, no ground truth); CT-SLAM cloud is the raw point cloud of [3]; patch point and surfel counts in sequence field",{"slug":337,"group":338,"sourceId":5,"sourceLabel":6,"table":339,"selfRows":164,"metrics":340,"seqs":347,"entrants":351,"cells":356,"outcomes":364,"locators":365,"hardware":366,"wordings":367,"notes":368},"elasticlidarfusion2018-table-i","elasticlidarfusion2018:Table I","Table I",[341,344],{"label":342,"unit":343,"statistic":95,"alignment":74},"No. State (optimisation state dimension)","count",{"label":345,"unit":346,"statistic":95,"alignment":74},"Elapsed Time (sec) of global loop-closure optimisation","s",[348],{"dataset":106,"sequence":349,"environment":350},"Fig. 1 office map","indoor office",[352,354],{"name":353,"methodId":5,"linkable":148,"proposed":148,"self":148},"Proposed (Elastic LiDAR Fusion, deformation graph)",{"name":355,"methodId":310,"linkable":148,"proposed":76,"self":76},"CT-SLAM [3] (global batch trajectory optimisation)",[357,359,360,362],[158,158,158,358,160,158,160,160,158],192,[158,162,158,183,160,158,160,160,158],[162,158,158,361,160,158,160,160,158],3396,[162,162,158,363,160,158,160,160,158],195.4,[],[339],[],[],[369],"global loop-closure optimisation cost for the map of Fig. 1 (office); proposed closes the loop at Fig. 5 (i), CT-SLAM batch-optimises the whole subsampled trajectory at the end (Fig. 5 (ii)); each state is 6-DoF; hardware not reported",[],1790510658114]