[{"data":1,"prerenderedAt":114},["ShallowReactive",2],{"method-rgbdslamv2_2014":3},{"method":4,"reference":58,"equipment":82,"figures":113,"results":71},{"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":22,"limitations":28,"sensors":34,"platform":36,"estimator":38,"association":39,"timeModel":40,"deskew":41,"loopClosure":42,"globalOptimization":43,"mapRepresentation":44,"prior":45,"outputGeometry":46,"compute":47,"codeUrl":48,"codeLicense":49,"relatedVersions":50},"rgbdslamv2_2014","Endres et al., 2014","RGBDSLAMv2","3-D Mapping With an RGB-D Camera",2014,"classic","C08","full_slam_with_global_correction","RGBDSLAMv2 只用 RGB-D 相機建立三維地圖。前端從彩色影像擷取 SIFT、SURF 或 ORB 特徵，以深度影像取得三維位置，再用 RANSAC 估計影格間的剛體轉換；候選影格包含前幾張影格、位姿圖上測地鄰域的抽樣以及關鍵影格，用來尋找迴圈閉合。作者提出以光束模型檢查深度影像間的自由空間衝突（EMM），剔除不可信的轉換，後端以 g2o 最佳化位姿圖並刪除誤差過大的邊。最後依軌跡把量測投影成點雲，或以 OctoMap 產生佔據體素地圖。","RGBDSLAMv2 builds 3-D maps from an RGB-D camera alone using keypoint matching with RANSAC, geodesic-neighbourhood and keyframe loop-closure candidates, a beam-based environment measurement model to reject wrong transforms, and g2o pose-graph optimisation with edge pruning, exporting point clouds or OctoMaps.","full_text_reviewed","peer_reviewed_published","supplementary","論文未在營建工地測試；評估使用 TUM RGB-D 資料集（辦公室尺度場景與工業廠房內的 Pioneer 機器人序列）以及 MIT Stata Center 序列。系統可直接輸出配準後點雲或 OctoMap，且作者指出結構光感測器在日光下無法使用、重複結構易造成錯誤配對，這些限制對室內裝修或機電階段的 RGB-D 掃描有參考價值（推論）。",[20,21],"public_benchmark","independent_reference",[23,24,25,26,27],"ATE RMSE 0.026 m on fr1\u002Fdesk, 0.087 m on fr1\u002Froom and 0.057 m on fr2\u002Fdesk; better than the best results reported for Kintinuous except fr2\u002Flarge no loop (Table I)","The EMM substantially reduces error on the challenging Robot SLAM (Pioneer) sequences, especially together with edge pruning (Figs. 9-10; Sec. IV-D)","Geodesic-neighbourhood sampling reduced the average error on Robot SLAM by 26% (Fig. 5 caption)","SIFT on GPU gives the highest accuracy (median RMSE 0.04 m on fr1) (Sec. IV-B)","OctoMap output at 2 cm is 4.2 to 25 MB versus 2 to 5 GB for unfiltered point clouds (Sec. III-F)",[29,30,31,32,33],"Repetitive structures (same chairs, windows, wallpaper, poles) cause systematic misassociation and bogus transforms that can distort the optimised graph (Secs. III-B, III-E)","Structured-light sensors are generally not usable in direct sunlight and have short range; feature-poor stretches with no depth occur in large halls (Secs. II, IV-A)","ORB or Shi-Tomasi plus SURF reach real-time rates but with an average error of about 15 cm on fr1 (Sec. IV-B)","OctoMap raycasting is expensive (about 1 s per 100 000 points at 5 cm; about 25 s per frame at 5 mm) and cannot be updated efficiently after large loop closures, so the map must be recreated (Sec. III-F)","The chosen metric (ATE) does not directly assess map quality (Sec. IV-A)",[35],"RGB-D camera (structured light: Microsoft Kinect, Asus Xtion Pro Live)",[37],"TUM RGB-D fr1 and fr2 sequences (the paper does not state how the camera was carried in these sequences)","pose-graph SLAM: pairwise 6-DOF transforms from 3-D feature correspondences via RANSAC with least-squares motion estimation, optional two-frame g2o refinement; global g2o optimisation (CSparse offline, PCG suggested online) with pruning of edges whose error remains high after convergence (Secs. III-B, III-E, IV-C)","sparse visual keypoints (SIFT on GPU, SURF, ORB or Shi-Tomasi plus SURF) matched by nearest to second-nearest ratio (Euclidean, Hellinger or Hamming distance), RANSAC with Mahalanobis inlier test; transforms validated by a beam-based environment measurement model (EMM) on subsampled depth images (Secs. III-B, III-C, IV-D)","discrete frames; frame-to-frame transforms between RGB-D images","not_applicable","candidate frames from n immediate predecessors, k frames sampled from the geodesic neighbourhood in the pose graph and l frames sampled from keyframes; validated by RANSAC and the EMM (Sec. III-D)","g2o pose-graph optimisation with Mahalanobis-based edge pruning (Secs. III-E, IV-C)","globally registered point cloud (optionally surfels) or OctoMap 3-D occupancy grid; paper recommends OctoMap for memory and free-space representation (Sec. III-F)","none (no odometry or other sensors used, even where wheel odometry was available)","optimised camera trajectory plus point cloud created by projecting the original depth measurements, or a textured OctoMap voxel occupancy map (2 cm maps of 4.2 to 25 MB versus 2 to 5 GB for unfiltered point clouds) (Sec. III-F)","Intel Core i7 3.40 GHz with nVidia GeForce GTX 570 (SIFT on GPU); offline processing of every frame at 5.04 to 15.2 Hz in Table I; median 13.0 Hz (9.1 to 16.4 Hz) on fr1 with SIFT (Fig. 8 caption); EMM 0.82 ms per bidirectional check (Sec. IV-D)","https:\u002F\u002Fgithub.com\u002Ffelixendres\u002Frgbdslam_v2","GPL-3.0 (COPYING file; README states GPL v.3)",[51,55],{"relation":52,"title":53,"doi_or_url":54},"conference_version","An evaluation of the RGB-D SLAM system (ICRA 2012; Endres, Hess, Engelhard, Sturm, Cremers, Burgard); reference [24] of the T-RO paper matches this Crossref record (title, authors, ICRA 2012, pp. 1691-1696)","10.1109\u002FICRA.2012.6225199",{"relation":56,"title":57,"doi_or_url":48},"code_release","felixendres\u002Frgbdslam_v2 (ROS package; README says it creates 3D point clouds or OctoMaps and cites this T-RO article)",{"id":5,"kind":59,"shortName":7,"title":8,"authors":60,"year":9,"venue":66,"venueType":67,"publisher":68,"volumeIssuePages":69,"doi":70,"arxivId":71,"url":72,"firstPublicDate":73,"publicationStatus":16,"metadataStatus":74,"fulltextStatus":15,"era":10,"classicReason":75,"codeUrl":48,"cluster":11,"topics":76,"mdpi":77,"verification":78,"label":6,"fulltextRoute":79,"versionRead":80,"addedByCensus":81},"method",[61,62,63,64,65],"Felix Endres","Jürgen Hess","Jürgen Sturm","Daniel Cremers","Wolfram Burgard","IEEE Transactions on Robotics","journal","IEEE","30(1):177-187","10.1109\u002Ftro.2013.2279412",null,"https:\u002F\u002Fieeexplore.ieee.org\u002Fdocument\u002F6594910","2013-09-09","metadata_verified","reproducible baseline: early open-source RGB-D SLAM (feature matching, RANSAC, g2o pose graph) whose ROS code exports registered point clouds or OctoMaps; it adds a beam-based environment measurement model to reject bad frame-to-frame transforms.",[11],false,"corrected","NTU institutional (Chrome)","version of record (IEEE Xplore HTML full text, T-RO 30(1), 2014)",true,[83,90,93,99,104,110],{"category":84,"model":85,"canonical":85,"role":86,"dataset":87,"specs":88,"locator":89},"rgbd","Microsoft Kinect","dataset sensor","TUM RGB-D benchmark","structured light; two Kinect units used in the benchmark","Sec. IV-A",{"category":84,"model":91,"canonical":91,"role":86,"dataset":87,"specs":92,"locator":89},"Asus Xtion Pro Live","structured light",{"category":94,"model":95,"canonical":95,"role":96,"dataset":87,"specs":97,"locator":98},"other","high-precision motion capturing system (model not reported)","reference or ground truth","synchronised ground-truth sensor trajectory","Secs. II, IV-A",{"category":100,"model":101,"canonical":101,"role":86,"dataset":102,"specs":103,"locator":89},"platform","Pioneer 3","TUM RGB-D Robot SLAM sequences","Kinect mounted on the robot; wheel odometry available but not used",{"category":105,"model":106,"canonical":106,"role":107,"dataset":71,"specs":108,"locator":109},"compute","Intel Core i7","compute for runtime","3.40 GHz; used for all experiments","Secs. IV, IV-B",{"category":105,"model":111,"canonical":111,"role":107,"dataset":71,"specs":112,"locator":109},"nVidia GeForce GTX 570","graphics card used for all experiments; SIFT computed on the GPU (SiftGPU)",[],1790510656090]