[{"data":1,"prerenderedAt":456},["ShallowReactive",2],{"method-dliom2023":3},{"method":4,"reference":66,"equipment":89,"figures":143,"results":144},{"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":35,"platform":38,"estimator":42,"association":43,"timeModel":44,"deskew":45,"loopClosure":46,"globalOptimization":47,"mapRepresentation":48,"prior":49,"outputGeometry":50,"compute":51,"codeUrl":52,"codeLicense":53,"relatedVersions":54},"dliom2023","Wang et al., 2023b","D-LIOM","D-LIOM: Tightly-Coupled Direct LiDAR-Inertial Odometry and Mapping",2023,"recent","C05","full_slam_with_global_correction","D-LIOM 把 Cartographer 式的直接配準改為與 IMU 緊耦合的 3D 版本：每個去畸變掃描不擷取特徵，直接以高斯牛頓法對齊到 3D 佔據機率子地圖，得到的位姿作為一元 LiDAR 因子，與 IMU 預積分及線上估計的重力先驗因子組成子地圖時間窗內的局部因子圖，同時更新 IMU 偏差並抑制橫滾與俯仰漂移。後端利用重力對齊，把 3D 子地圖投影成 2D 影像，以 SURF 特徵與 FLANN 比對偵測迴圈並由 RANSAC 求得平面位移與旋轉，再以分支定界搜尋高度差、做精細配準後最佳化全域位姿圖。系統也支援多 LiDAR 輸入與 6 軸 IMU 的靜態或動態初始化。","Tightly coupled direct LIO and mapping that registers raw points to Cartographer-style 3D probability submaps, fuses the resulting pose with IMU preintegration and an online gravity prior in a submap-window factor graph, supports multiple LiDARs and dynamic 6-axis IMU initialization, and closes loops by SURF matching of gravity-aligned 2D projections of submaps followed by 3D refinement.","full_text_reviewed","peer_reviewed_published","supplementary","論文未在營建場域測試；資料為同濟校園手持（騎車與步行）、NTU VIRAL 無人機（作者稱為室內資料）與 Complex Urban 城市車載序列。Ghadimzadeh Alamdari 等人 [ghadimzadeh2025slamnde] 在針對基礎設施非破壞檢測的比較中，將 D-LIOM 列為因程式庫不一致而未能執行。其直接機率子地圖配準不依賴特徵，並支援多 LiDAR 與傾斜安裝，理論上適合特徵稀少或安裝角度受限的工地平台，但實際可重現性受程式庫狀態限制（推論）。",[20,21,22],"public_benchmark","independent_reference","cross_site",[24,25,26,27,28],"With both VIRAL LiDARs, weighted average positioning error 0.10 m versus 0.15 m for LIO-SAM and 1.82 m for LIOM (Table III)","Revisiting errors below 1 m on four 3 to 3.5 km TONGJI sequences where Carto3D, LOAM, LIOM and LIO-SAM had errors from 1.08 m up to hundreds of metres or failed (Table IV)","Submap-to-submap loop detection reached 90.94% precision and 95.45% recall (Table VII)","Ran on Complex Urban Urban-09 and Urban-10 with strongly inclined LiDARs where the compared methods produced no meaningful results (Sec. IV-B-4)","Sharper reconstructions than LIO-SAM under rotations above 223 deg\u002Fs (Sec. IV-B-2, Fig. 5)",[30,31,32,33,34],"With a single horizontal LiDAR on VIRAL it was slightly less accurate than LIO-SAM, which the authors relate to submap resolution in small, slow scenes (Table III, Sec. IV-C-1)","Loop detection can fail through feature mismatches and similar structures (false negatives and false positives) (Sec. IV-E, Fig. 9)","The time offset between IMU and LiDAR is ignored (Sec. IV-F)","TONGJI revisiting-error reference comes from NDT registration of revisited scans, not independent ground truth; Complex Urban GPS is only an approximate reference (Sec. IV-B-4, IV-C-2)","The published Sec. III-D contained errors later corrected by the authors in the repository (README)",[36,37],"one or more 3D spinning LiDARs (16-line RoboSense in the authors' device; two Ouster OS1-16 in NTU VIRAL; two inclined 16-line Velodyne in Complex Urban)","6-axis IMU (built-in consumer-grade IMU at 400 Hz; VectorNav VN100 in VIRAL)",[39,40,41],"handheld (authors' device, carried while cycling and walking)","UAV (NTU VIRAL, DJI M600)","vehicle (Complex Urban Dataset)","front-end local factor graph in GTSAM over the time window of the current submap with LiDAR odometry unary factors from direct scan-to-submap matching, IMU preintegration factors and an online gravity-prior factor on roll and pitch; back-end global sparse pose graph in Ceres (Sec. III-E, IV-A)","direct: raw deskewed points are registered to a 3D probability (occupancy log-odds) submap by Gauss-Newton maximization of voxel probability using map gradients; no feature extraction (Sec. III-E-2)","discrete scan nodes with IMU preintegration; LiDAR-IMU time offset ignored (Sec. III-C, IV-F)","IMU preintegration between scans deskews each point to the previous body frame; points of auxiliary LiDARs are merged by timestamp with the primary LiDAR (Sec. III-C-2)","yes; gravity-aligned 3D submaps projected to 2D images, SURF keypoints matched with FLANN, RANSAC 3-DoF transform, then branch-and-bound search of the vertical offset and fine scan-to-submap registration (Sec. III-F)","global sparse pose graph over submaps and nodes with loop constraints (Sec. III-F)","3D probability submaps stored as octrees of occupancy log-odds voxels (Cartographer-style), plus 2D projections for loop detection (Sec. III-E-2)","LiDAR-IMU extrinsics calibrated offline; gravity magnitude assumed known (Sec. III-D, IV-A)","trajectory and probability submaps; dense point cloud maps shown qualitatively (Figs. 5-6)","real time on a workstation with two Intel Xeon E5-2620V3 at 2.4 GHz and 32 GB RAM: 27.5 ms odometry and 33.5 ms mapping per 16-line scan, 54.5 and 60.1 ms per 64-line scan (Table V, Sec. IV-A)","https:\u002F\u002Fgithub.com\u002FpeterWon\u002FD-LIOM","not_stated (no LICENSE file found; builds on Cartographer)",[55,59,62],{"relation":56,"title":57,"doi_or_url":58},"correction_statement_in_article","Author-revised version in the code repository correcting typos and errors in Sec. III-D (system initialization) of the published article","https:\u002F\u002Fgithub.com\u002FpeterWon\u002FD-LIOM\u002Fblob\u002Fmaster\u002Fpaper\u002FTMM_D_LIOM_FINAL_VERSION.pdf",{"relation":60,"title":61,"doi_or_url":52},"code_release","peterWon\u002FD-LIOM (built on Cartographer; TONGJI demo data)",{"relation":63,"title":64,"doi_or_url":65},"project_page","D-LIOM project page named in the paper footnote (not opened)","https:\u002F\u002Fcslinzhang.github.io\u002FD-LIOM\u002FD-LIOM.html",{"id":5,"kind":67,"shortName":7,"title":8,"authors":68,"year":9,"venue":73,"venueType":74,"publisher":75,"volumeIssuePages":76,"doi":77,"arxivId":78,"url":79,"firstPublicDate":80,"publicationStatus":16,"metadataStatus":81,"fulltextStatus":15,"era":10,"classicReason":82,"codeUrl":52,"cluster":11,"topics":83,"mdpi":84,"verification":85,"label":6,"fulltextRoute":86,"versionRead":87,"addedByCensus":88},"method",[69,70,71,72],"Zhong Wang","Lin Zhang","Ying Shen","Yicong Zhou","IEEE Transactions on Multimedia","journal","IEEE","25:3905-3920","10.1109\u002Ftmm.2022.3168423",null,"https:\u002F\u002Fdoi.org\u002F10.1109\u002FTMM.2022.3168423","2022-04-19","metadata_verified","not_applicable",[11],false,"corrected","author copy","Author-revised final version from the code repository (PDF created 2022-06-30), which corrects Sec. III-D of the published article; IEEE version of record not read",true,[90,97,101,106,112,116,120,125,127,133,139],{"category":91,"model":92,"canonical":92,"role":93,"dataset":94,"specs":95,"locator":96},"lidar","RoboSense 16-line LiDAR (model not stated)","method input","TONGJI dataset (authors' handheld device)","10 Hz","Sec. IV-A-2; Fig. 4(a)",{"category":98,"model":99,"canonical":99,"role":93,"dataset":94,"specs":100,"locator":96},"imu","built-in consumer-grade IMU of a ZED camera (model not stated)","400 Hz; noise statistics from Woodman's approach",{"category":102,"model":103,"canonical":103,"role":93,"dataset":94,"specs":104,"locator":105},"platform","self-developed handheld device","carried while cycling (TJ-1 to TJ-4) and walking (TJ-5); aggressive rotations up to 232 deg\u002Fs (TJ-6, TJ-7)","Sec. IV-A-2; Table I",{"category":102,"model":107,"canonical":107,"role":108,"dataset":109,"specs":110,"locator":111},"DJI M600 UAV","dataset sensor","NTU VIRAL","not_reported","Sec. IV-A-2; Fig. 4(b)",{"category":91,"model":113,"canonical":114,"role":108,"dataset":109,"specs":110,"locator":115},"Ouster OS1-16-gen-1 (two units, horizontal and vertical)","Ouster OS1-16","Sec. IV-A-2; Sec. IV-C-1",{"category":98,"model":117,"canonical":118,"role":108,"dataset":109,"specs":110,"locator":119},"VectorNav-VN100","VectorNav VN100","Sec. IV-A-2",{"category":91,"model":121,"canonical":121,"role":108,"dataset":122,"specs":123,"locator":124},"two 16-line Velodyne LiDARs (inclined)","Complex Urban Dataset","large tilt angles","Sec. IV-A-2; Sec. IV-B-4",{"category":98,"model":126,"canonical":126,"role":108,"dataset":122,"specs":110,"locator":119},"AHRS IMU (model not stated)",{"category":128,"model":129,"canonical":129,"role":130,"dataset":122,"specs":131,"locator":132},"gnss","GPS (model not stated)","reference or ground truth","unstable under high-rise buildings; approximate reference only","Sec. IV-B-4",{"category":134,"model":135,"canonical":135,"role":136,"dataset":78,"specs":137,"locator":138},"compute","two Intel Xeon E5-2620V3","compute for runtime","2.4 GHz, 32 GB RAM workstation","Sec. IV-A-1",{"category":102,"model":140,"canonical":140,"role":108,"dataset":122,"specs":141,"locator":142},"car (Complex Urban vehicle-mounted platform)","two inclined LiDARs and an IMU; each sequence about 1 hour and nearly 2 square kilometres","Sec. IV-A-2; Sec. IV-B-4; Fig. 4(c)",[],{"totalRows":145,"groupCount":146,"groups":147,"others":450},42,5,[148,268,323,388],{"slug":149,"group":150,"sourceId":5,"sourceLabel":6,"table":151,"selfRows":152,"metrics":153,"seqs":157,"entrants":179,"cells":190,"outcomes":262,"locators":263,"hardware":264,"wordings":265,"notes":266},"dliom2023-table-iii","dliom2023:Table III","Table III",20,[154],{"label":155,"unit":156,"statistic":110,"alignment":110},"absolute positioning error","m",[158,161,163,165,167,169,171,173,175,177],{"dataset":109,"sequence":159,"environment":160},"eee01","UAV; described by the authors as an indoor dataset collected with slow motion (Sec. I, IV-C-1)",{"dataset":109,"sequence":162,"environment":160},"eee02",{"dataset":109,"sequence":164,"environment":160},"eee03",{"dataset":109,"sequence":166,"environment":160},"nya01",{"dataset":109,"sequence":168,"environment":160},"nya02",{"dataset":109,"sequence":170,"environment":160},"nya03",{"dataset":109,"sequence":172,"environment":160},"sbs01",{"dataset":109,"sequence":174,"environment":160},"sbs02",{"dataset":109,"sequence":176,"environment":160},"sbs03",{"dataset":109,"sequence":178,"environment":160},"w-avg",[180,183,186,188],{"name":181,"methodId":182,"linkable":88,"proposed":84,"self":84},"LIOM [9]","liomapping2019",{"name":184,"methodId":185,"linkable":88,"proposed":84,"self":84},"LIO-SAM [10]","liosam2020",{"name":187,"methodId":5,"linkable":88,"proposed":88,"self":88},"Ours (H)",{"name":189,"methodId":5,"linkable":88,"proposed":88,"self":88},"Ours (HV)",[191,195,198,201,204,207,208,211,214,216,219,221,223,225,227,228,230,232,233,234,236,238,240,241,243,244,245,247,248,250,252,253,254,255,256,257,258,259,260,261],[192,192,192,193,194,192,194,194,192],0,1.06,-1,[192,192,196,197,194,192,194,194,192],1,0.72,[192,192,199,200,194,192,194,194,192],2,1.03,[192,192,202,203,194,192,194,194,192],3,2.24,[192,192,205,206,194,192,194,194,192],4,1.97,[192,192,146,202,194,192,194,194,192],[192,192,209,210,194,192,194,194,192],6,1.67,[192,192,212,213,194,192,194,194,192],7,1.81,[192,192,215,199,194,192,194,194,192],8,[192,192,217,218,194,192,194,194,192],9,1.82,[196,192,192,220,194,192,194,194,192],0.1,[196,192,196,222,194,192,194,194,192],0.08,[196,192,199,224,194,192,194,194,192],0.12,[196,192,202,226,194,192,194,194,192],0.09,[196,192,205,224,194,192,194,194,192],[196,192,146,229,194,192,194,194,192],0.42,[196,192,209,231,194,192,194,194,192],0.11,[196,192,212,224,194,192,194,194,192],[196,192,215,220,194,192,194,194,192],[196,192,217,235,194,192,194,194,192],0.15,[199,192,192,237,194,192,194,194,192],0.23,[199,192,196,239,194,192,194,194,192],0.24,[199,192,199,231,194,192,194,194,192],[199,192,202,242,194,192,194,194,192],0.14,[199,192,205,242,194,192,194,194,192],[199,192,146,235,194,192,194,194,192],[199,192,209,246,194,192,194,194,192],0.16,[199,192,212,235,194,192,194,194,192],[199,192,215,249,194,192,194,194,192],0.19,[199,192,217,251,194,192,194,194,192],0.17,[202,192,192,226,194,192,194,194,192],[202,192,196,222,194,192,194,194,192],[202,192,199,220,194,192,194,194,192],[202,192,202,220,194,192,194,194,192],[202,192,205,220,194,192,194,194,192],[202,192,146,220,194,192,194,194,192],[202,192,209,224,194,192,194,194,192],[202,192,212,231,194,192,194,194,192],[202,192,215,235,194,192,194,194,192],[202,192,217,220,194,192,194,194,192],[],[151],[],[],[267],"NTU VIRAL absolute positioning errors; LIOM, LIO-SAM and Ours (H) use the horizontal LiDAR only, Ours (HV) uses horizontal and vertical LiDARs; w-avg weighted by number of poses",{"slug":269,"group":270,"sourceId":5,"sourceLabel":6,"table":271,"selfRows":272,"metrics":273,"seqs":277,"entrants":288,"cells":294,"outcomes":317,"locators":318,"hardware":319,"wordings":320,"notes":321},"dliom2023-table-vi","dliom2023:Table VI","Table VI",12,[274],{"label":275,"unit":156,"statistic":110,"alignment":276},"revisiting error","none",[278,282,284,286],{"dataset":279,"sequence":280,"environment":281},"TONGJI dataset","TJ-1 (3014.68 m)","Tongji campus, handheld while cycling",{"dataset":279,"sequence":283,"environment":281},"TJ-2 (3004.97 m)",{"dataset":279,"sequence":285,"environment":281},"TJ-3 (3170.73 m)",{"dataset":279,"sequence":287,"environment":281},"TJ-4 (3500.25 m)",[289,291,293],{"name":290,"methodId":5,"linkable":88,"proposed":84,"self":88},"D-LIOM WoG (without gravity factor)",{"name":292,"methodId":5,"linkable":88,"proposed":84,"self":88},"D-LIOM WoL (without loop detection)",{"name":7,"methodId":5,"linkable":88,"proposed":88,"self":88},[295,297,299,301,303,305,307,309,311,313,314,315],[192,192,192,296,194,192,194,194,192],7.14,[192,192,196,298,194,192,194,194,192],0.65,[192,192,199,300,194,192,194,194,192],0.59,[192,192,202,302,194,192,194,194,192],11.45,[196,192,192,304,194,192,194,194,192],12.14,[196,192,196,306,194,192,194,194,192],19.81,[196,192,199,308,194,192,194,194,192],3.68,[196,192,202,310,194,192,194,194,192],61.37,[199,192,192,312,194,192,194,194,192],0.81,[199,192,196,298,194,192,194,194,192],[199,192,199,300,194,192,194,194,192],[199,192,202,316,194,192,194,194,192],0.58,[],[271],[],[],[322],"Ablation on TONGJI: D-LIOM without gravity factor (WoG) and without submap-to-submap loop detection (WoL); revisiting error",{"slug":324,"group":325,"sourceId":5,"sourceLabel":6,"table":326,"selfRows":205,"metrics":327,"seqs":329,"entrants":335,"cells":347,"outcomes":381,"locators":383,"hardware":384,"wordings":385,"notes":386},"dliom2023-table-iv","dliom2023:Table IV","Table IV",[328],{"label":275,"unit":156,"statistic":110,"alignment":276},[330,332,333,334],{"dataset":279,"sequence":280,"environment":331},"Tongji campus structured and unstructured areas, handheld while cycling",{"dataset":279,"sequence":283,"environment":331},{"dataset":279,"sequence":285,"environment":331},{"dataset":279,"sequence":287,"environment":331},[336,339,342,344,346],{"name":337,"methodId":338,"linkable":88,"proposed":84,"self":84},"Carto3D","cartographer2016",{"name":340,"methodId":341,"linkable":88,"proposed":84,"self":84},"LOAM","loam2014",{"name":343,"methodId":182,"linkable":88,"proposed":84,"self":84},"LIOM",{"name":345,"methodId":185,"linkable":88,"proposed":84,"self":84},"LIO-SAM",{"name":7,"methodId":5,"linkable":88,"proposed":88,"self":88},[348,350,352,354,356,358,360,362,364,365,366,368,369,371,373,375,377,378,379,380],[192,192,192,349,194,192,194,194,192],14.7,[192,192,196,351,194,192,194,194,192],39.97,[192,192,199,353,194,192,194,194,192],34.89,[192,192,202,355,194,192,194,194,192],52.76,[196,192,192,357,194,192,194,194,192],90.33,[196,192,196,359,194,192,194,194,192],305.76,[196,192,199,361,194,192,194,194,192],129.8,[196,192,202,363,194,192,194,194,192],380.1,[199,192,192,78,192,192,194,194,192],[199,192,196,78,192,192,194,194,192],[199,192,199,367,194,192,194,194,192],22.81,[199,192,202,78,192,192,194,194,192],[202,192,192,370,194,192,194,194,192],1.08,[202,192,196,372,194,192,194,194,192],27.87,[202,192,199,374,194,192,194,194,192],14.16,[202,192,202,376,194,192,194,194,192],40.21,[205,192,192,312,194,192,194,194,192],[205,192,196,298,194,192,194,194,192],[205,192,199,300,194,192,194,194,192],[205,192,202,316,194,192,194,194,192],[382],"failed",[326],[],[],[387],"TONGJI self-collected handheld data; revisiting error = difference between each method's relative pose of two revisiting nodes and the relative pose from NDT registration of the revisited scans",{"slug":389,"group":390,"sourceId":5,"sourceLabel":6,"table":391,"selfRows":205,"metrics":392,"seqs":399,"entrants":405,"cells":410,"outcomes":443,"locators":444,"hardware":445,"wordings":447,"notes":448},"dliom2023-table-v","dliom2023:Table V","Table V",[393,397],{"label":394,"unit":395,"statistic":396,"alignment":82},"Odometry time cost","ms","mean",{"label":398,"unit":395,"statistic":396,"alignment":82},"Mapping time cost",[400,403],{"dataset":110,"sequence":401,"environment":402},"16 lines LiDAR scan","per LiDAR scan",{"dataset":110,"sequence":404,"environment":402},"64 lines LiDAR scan",[406,407,408,409],{"name":340,"methodId":341,"linkable":88,"proposed":84,"self":84},{"name":343,"methodId":182,"linkable":88,"proposed":84,"self":84},{"name":345,"methodId":185,"linkable":88,"proposed":84,"self":84},{"name":7,"methodId":5,"linkable":88,"proposed":88,"self":88},[411,413,415,417,419,421,423,425,427,429,431,433,435,437,439,441],[192,192,192,412,194,192,192,194,192],55.6,[196,192,192,414,194,192,192,194,192],684.7,[199,192,192,416,194,192,192,194,192],25.9,[202,192,192,418,194,192,192,194,192],27.5,[192,196,192,420,194,192,192,194,192],166.6,[196,196,192,422,194,192,192,194,192],140.6,[199,196,192,424,194,192,192,194,192],124.1,[202,196,192,426,194,192,192,194,192],33.5,[192,192,196,428,194,192,192,194,192],116.2,[196,192,196,430,194,192,192,194,192],1181.1,[199,192,196,432,194,192,192,194,192],47.2,[202,192,196,434,194,192,192,194,192],54.5,[192,196,196,436,194,192,192,194,192],117.7,[196,196,196,438,194,192,192,194,192],236.9,[199,196,196,440,194,192,192,194,192],158,[202,196,196,442,194,192,192,194,192],60.1,[],[391],[446],"two Intel Xeon E5-2620V3 2.4 GHz, 32 GB RAM",[],[449],"Time cost per scan of the Odometry and Mapping modules for 16-line and 64-line LiDAR input; D-LIOM back-end time summed and averaged per frame",[451],{"group":452,"slug":453,"sourceLabel":6,"table":454,"selfRows":199,"datasets":455},"dliom2023:Table VII","dliom2023-table-vii","Table VII",[279],1790510657791]