[{"data":1,"prerenderedAt":161},["ShallowReactive",2],{"method-dpslam2003":3},{"method":4,"reference":56,"equipment":76,"figures":93,"results":94},{"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":26,"sensors":32,"platform":35,"estimator":37,"association":38,"timeModel":39,"deskew":40,"loopClosure":41,"globalOptimization":42,"mapRepresentation":43,"prior":42,"outputGeometry":44,"compute":45,"codeUrl":46,"codeLicense":47,"relatedVersions":48},"dpslam2003","Eliazar & Parr, 2003","DP-SLAM","DP-SLAM: Fast, Robust Simultaneous Localization and Mapping Without Predetermined Landmarks",2003,"classic","C01","full_slam_with_global_correction","DP-SLAM 以粒子濾波同時追蹤機器人位姿與地圖假設，不需預先指定地標。為避免每個粒子複製整張地圖，作者提出分散式粒子建圖（DP-mapping）：全系統只保留一張佔據網格，每格以平衡樹記錄曾更新該格的粒子 ID，並維護經修剪與合併的最小粒子祖先樹；粒子查詢某格時，沿祖先找出最近一次的觀測紀錄。如此可同時維持數千張候選地圖，並在沒有明確迴圈閉合步驟下閉合約 60 m 的迴圈。","Laser-only particle filter over poses and maps that shares one occupancy grid among particles through an ancestry tree and per-cell balanced trees (DP-mapping), keeping thousands of map hypotheses and closing loops without explicit loop closing.","full_text_reviewed","peer_reviewed_published","background","未在營建場域驗證；測試為 Duke 大學資訊系館二樓約 16 m 乘 14 m 的走廊迴圈，只以地圖目視比較，沒有定量誤差。其保存多張地圖假設、待證據充足再收斂的做法，與營建場域中暫時結構造成量測歧義時的建圖穩健性相關（推論）。",[20],"completed_building",[22,23,24,25],"Closed a 60 m loop with 9000 particles and no discernible misalignment, whereas keeping only the best particle's single map produced a considerable misalignment (Sec. 4.2; Figs. 1-2).","Ancestry trees coalesced as recently as about 20 generations and never beyond 90 in the test domain, keeping storage close to linear in practice (Sec. 4.3).","Particle culling with k = 6 gave about a 6x speed-up, and run time was close to data-collection time (Sec. 4).","No assumptions about the environment or the existence of loops (Sec. 4).",[27,28,29,30,31],"The binary occupancy grid causes discretization errors at object edges and confusion with small objects such as power cords hanging from desks (Sec. 5).","The laser is treated as deterministic when updating the map but as noisy when localizing (Sec. 5).","A fixed-height laser gives an incomplete view of the environment (Sec. 5).","Sparse-data environments may cause premature coalescence unless an excessive number of particles is used (Sec. 4.3).","Using one in four laser casts required 30,000 particles for good maps (Sec. 4.3).",[33,34],"2D laser range finder (SICK)","wheel odometry (shaft encoders)",[36],"wheeled UGV (iRobot ATRV Jr., skid steering)","particle filter over robot poses and maps with a calibrated odometry motion model; particle culling evaluates the posterior in k passes over disjoint subsets of laser readings and drops poor particles early (Sec. 2, 4)","ray tracing each laser cast through the particle's map to the first obstruction; Gaussian discrepancy with 5 cm standard deviation (Sec. 2.1)","discrete poses (new observation after about 20 cm of motion)","not_reported","implicit through maintaining multiple map hypotheses; no explicit loop-closing step or environment assumption (Sec. 4)","none","single binary occupancy grid (3 cm cells) in which each cell stores a balanced tree keyed by the IDs of particles that updated it, plus a pruned and collapsed minimal particle ancestry tree (Sec. 3.2; Sec. 4)","2D occupancy grid (cross-section at the 7 cm laser height)","2.4 GHz Pentium 4: run time close to data-collection time; culling with k = 6 gave about a 6x speed-up; worst-case cost O(ADP lg P) per sweep (Sec. 3.3-3.4; Sec. 4)","https:\u002F\u002Fgithub.com\u002FOpenSLAM-org\u002Fopenslam_dpslam","Duke University research licence stated on the OpenSLAM.org DP-SLAM page (provided without cost and as is, with an indemnification clause); not an OSI-approved licence",[49,52],{"relation":50,"title":51,"doi_or_url":46},"code_release","OpenSLAM DP-SLAM source",{"relation":53,"title":54,"doi_or_url":55},"follow_up","DP-SLAM 2.0 (listed on the OpenSLAM page; not read)","https:\u002F\u002Fopenslam-org.github.io\u002Fdpslam.html",{"id":5,"kind":57,"shortName":7,"title":8,"authors":58,"year":9,"venue":61,"venueType":62,"publisher":63,"volumeIssuePages":64,"doi":65,"arxivId":65,"url":66,"firstPublicDate":67,"publicationStatus":16,"metadataStatus":68,"fulltextStatus":15,"era":10,"classicReason":69,"codeUrl":46,"cluster":11,"topics":70,"mdpi":71,"verification":72,"label":6,"fulltextRoute":73,"versionRead":74,"addedByCensus":75},"method",[59,60],"Austin Eliazar","Ronald Parr","Proceedings of the Eighteenth International Joint Conference on Artificial Intelligence (IJCAI-03)","conference","International Joint Conferences on Artificial Intelligence (IJCAI)","pp. 1135-1142",null,"https:\u002F\u002Fwww.ijcai.org\u002FProceedings\u002F03\u002FPapers\u002F163.pdf","2003","metadata_verified","principle reused: laser-only particle filter over both poses and full maps without landmarks, made tractable by sharing one occupancy grid among particles through ancestry trees (DP-mapping); open code on OpenSLAM. Its place between landmark FastSLAM [fastslam2002] and grid RBPF mappers such as GMapping [gmapping2007] is a reviewer synthesis.",[11],false,"corrected","publisher OA","IJCAI-03 proceedings PDF (pp. 1135-1142) from ijcai.org",true,[77,83,87],{"category":78,"model":79,"canonical":79,"role":80,"dataset":65,"specs":81,"locator":82},"platform","iRobot ATRV Jr.","method input","skid steering; shaft encoders unreliable when turning","Sec. 4.1",{"category":84,"model":85,"canonical":85,"role":80,"dataset":65,"specs":86,"locator":82},"lidar","SICK laser range finder","front-mounted 7 cm above the floor; 180 deg at 1 deg spacing; effective range up to 8 m; distance error typically below 5 mm",{"category":88,"model":89,"canonical":89,"role":90,"dataset":65,"specs":91,"locator":92},"compute","2.4 GHz Pentium 4","compute for runtime","fast PC used for offline processing of the logged data","Sec. 4",[],{"totalRows":95,"groupCount":96,"groups":97,"others":160},5,1,[98],{"slug":99,"group":100,"sourceId":5,"sourceLabel":6,"table":101,"selfRows":95,"metrics":102,"seqs":117,"entrants":124,"cells":130,"outcomes":146,"locators":150,"hardware":153,"wordings":154,"notes":155},"dpslam2003-text-sec-4","dpslam2003:Text Sec. 4","Text Sec. 4",[103,106,108,111,114],{"label":104,"unit":105,"statistic":40,"alignment":42},"number of particles for the map closing the loop with no discernible misalignment","particles",{"label":107,"unit":105,"statistic":40,"alignment":42},"particles required to produce good maps",{"label":109,"unit":110,"statistic":40,"alignment":42},"generations to the point of coalescence (often as recent as)","generations",{"label":112,"unit":110,"statistic":113,"alignment":42},"generations to the point of coalescence (never exceeded)","max",{"label":115,"unit":116,"statistic":40,"alignment":42},"speedup from particle culling (approximately)","x",[118,122],{"dataset":119,"sequence":120,"environment":121},"Duke University Computer Science building, 2nd floor (iRobot ATRV Jr., SICK)","hallway loop log","indoor building",{"dataset":123,"sequence":120,"environment":121},"Duke University Computer Science building, 2nd floor",[125,126,128],{"name":7,"methodId":5,"linkable":75,"proposed":75,"self":75},{"name":127,"methodId":5,"linkable":75,"proposed":71,"self":75},"DP-SLAM (handicapped, 1 in 4 laser casts)",{"name":129,"methodId":5,"linkable":75,"proposed":75,"self":75},"DP-SLAM with culling (k = 6)",[131,135,137,140,143],[132,132,132,133,132,132,132,134,132],0,9000,-1,[96,96,96,136,134,96,134,134,96],30000,[132,138,96,139,96,96,134,134,138],2,20,[132,141,96,142,134,96,134,134,138],3,90,[138,144,96,145,138,138,132,134,141],4,6,[147,148,149],"loop closed with no discernible misalignment","often as recent as 20 generations (wording of Sec. 4.3)","approximate",[151,152,92],"Sec. 4.2; Fig. 1","Sec. 4.3",[89],[],[156,157,158,159],"Hallway loop about 16 m x 14 m, 60 m traveled before re-observing the start; 3 cm grid","Handicapped DP-SLAM ignoring three of every four laser casts","Depth of coalescence of the ancestry tree over the run","Particle culling in k passes with k = 6",[],1790510661224]