[{"data":1,"prerenderedAt":309},["ShallowReactive",2],{"method-kazhdan2013screened":3},{"method":4,"reference":44,"equipment":64,"figures":71,"results":72},{"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":26,"sensors":29,"platform":30,"estimator":31,"association":31,"timeModel":31,"deskew":31,"loopClosure":31,"globalOptimization":32,"mapRepresentation":33,"prior":34,"outputGeometry":35,"compute":36,"codeUrl":37,"codeLicense":38,"relatedVersions":39},"kazhdan2013screened","Kazhdan & Hoppe, 2013","Screened Poisson Surface Reconstruction","Screened poisson surface reconstruction",2013,"classic","C12","map_representation_or_reconstruction","此版本在原泊松重建中加入點位置的軟約束（screening term），使重建等值面更貼近輸入點，以減輕原方法的過度平滑；因約束只定義在稀疏點集上，線性系統的稀疏結構不變，仍可用多重網格（multigrid）求解，並透過演算法改良使時間複雜度對點數呈線性。作者另支援 Neumann 邊界條件，讓缺資料區的曲面可延伸到定義域邊界，而非強制封閉。","Adds sparse point-interpolation (screening) constraints to Poisson reconstruction to reduce over-smoothing, keeps a multigrid-solvable sparse system with linear-time complexity, and supports Neumann boundaries.","full_text_reviewed","peer_reviewed_published","main_body","未在營建資料驗證，測試資料為雕像與機械零件等物件掃描及合成掃描。對 SLAM 點雲而言，作者報告的「掃描未對齊時表面起伏」代表軌跡或迴圈誤差會直接轉成網格表面瑕疵（推論）；作者也建議雜訊大時降低 screening 權重 α，而資料缺漏處的外插能力不如 SSD。",[20,21,22],"public_benchmark","simulation","controlled_experiment",[24,25],"[\"Reconstructions better capture input data at similar complexity and time as unscreened Poisson (Sec. 1, Fig. 1).\", \"Solver complexity reduced to linear in number of points (abstract).\", \"One-sided RMS error from held-out validation points is always lower than original Poisson and Wavelet and comparable to or lower than SSD on real scans","the same ordering holds on clean uniformly sampled data (Sec. 6.1, Figs. 4b and 5b).\", \"Faster than original Poisson at depths 9 to 11, e.g. David at depth 10: 182 s versus 412 s, while SSD needed 19,158 s (Table I).\", \"Authors state that with screening off (alpha = 0) the new solver is 2 to 3 times faster than the original Poisson implementation, of which a factor 1.1 to 1.6 comes from multithreading (Sec. 6.2).\"]",[27,28],"[\"With misaligned input scans, the screened reconstruction produces a pock-marked surface that undulates between scans (Sec. 6.3).\", \"For noisy data the screening weight must be reduced (Sec. 6.3).\", \"Neumann boundaries bias surfaces to cross the domain boundary orthogonally (Sec. 4.4, Fig. 3).\", \"SSD extrapolates better into regions of missing data (e.g. the Anchor model's cylindrical hole), giving SSD the smallest distance to ground truth in the Berger et al. benchmark (Sec. 6.1, Fig. 6","Sec. 7).\", \"Memory is higher than original Poisson, e.g. David at depth 10: 2194 MB versus 1498 MB (Table I).\", \"Point-to-surface RMS can favour the screened result even when misaligned scans make it visually worse (Sec. 6.3).\"]",[],[],"not_applicable","global screened Poisson solve with multigrid on an octree","implicit function on an adaptive octree with point-value (screening) constraints","oriented normals required","triangle mesh isosurface; Dirichlet (closed) or Neumann (open to domain boundary) boundary behavior","offline, CPU; C++ with OpenMP multithreading and conjugate-gradient relaxation at each multigrid level (Sec. 5.4); Table I timings on a laptop with a quad-core Intel Core i7 and 8 GB RAM (Sec. 6.2)","https:\u002F\u002Fgithub.com\u002Fmkazhdan\u002FPoissonRecon","MIT (repository LICENSE)",[40],{"relation":41,"title":42,"doi_or_url":43},"predecessor_method","Poisson Surface Reconstruction (Kazhdan, Bolitho and Hoppe, SGP 2006), the unscreened predecessor method; not a conference version of this article","10.2312\u002FSGP\u002FSGP06\u002F061-070",{"id":5,"kind":45,"shortName":7,"title":8,"authors":46,"year":9,"venue":49,"venueType":50,"publisher":51,"volumeIssuePages":52,"doi":53,"arxivId":54,"url":55,"firstPublicDate":56,"publicationStatus":16,"metadataStatus":57,"fulltextStatus":15,"era":10,"classicReason":58,"codeUrl":37,"cluster":11,"topics":59,"mdpi":60,"verification":61,"label":6,"fulltextRoute":62,"versionRead":63,"addedByCensus":60},"method",[47,48],"Michael Kazhdan","Hugues Hoppe","ACM Transactions on Graphics","journal","ACM","32(3):1-13","10.1145\u002F2487228.2487237",null,"https:\u002F\u002Fwww.cs.jhu.edu\u002F~misha\u002FMyPapers\u002FToG13.pdf","2013-06","metadata_verified","reproducible baseline: the screened variant (point interpolation constraints, Neumann boundaries, linear-time solver) is the implementation typically used as the offline Poisson baseline, e.g., the official code used in ImMesh experiments.",[11],false,"corrected","author copy","Author preprint PDF (cs.jhu.edu\u002F~misha\u002FMyPapers\u002FToG13.pdf, 13 pp., placeholder ACM reference 'Article XXX', 'Received May 2012; accepted December 2012'); ACM TOG 32(3) version of record not compared",[65],{"category":66,"model":67,"canonical":67,"role":68,"dataset":54,"specs":69,"locator":70},"compute","Intel Core i7","compute for runtime","quad-core; laptop with 8 GB RAM","Sec. 6.2",[],{"totalRows":73,"groupCount":74,"groups":75,"others":308},28,2,[76,271],{"slug":77,"group":78,"sourceId":5,"sourceLabel":6,"table":79,"selfRows":80,"metrics":81,"seqs":91,"entrants":107,"cells":120,"outcomes":263,"locators":265,"hardware":266,"wordings":268,"notes":269},"kazhdan2013screened-table-i","kazhdan2013screened:Table I","Table I",24,[82,86,88],{"label":83,"unit":84,"statistic":85,"alignment":85},"Time in seconds","s","not_reported",{"label":87,"unit":84,"statistic":85,"alignment":85},"Time in seconds (bracketed, alpha = 0)",{"label":89,"unit":90,"statistic":85,"alignment":85},"Memory in MB","MB",[92,96,98,100,102,104,105,106],{"dataset":93,"sequence":94,"environment":95},"Neptune (Aim@Shape)","depth 8","object scan",{"dataset":93,"sequence":97,"environment":95},"depth 9",{"dataset":93,"sequence":99,"environment":95},"depth 10",{"dataset":93,"sequence":101,"environment":95},"depth 11",{"dataset":103,"sequence":94,"environment":95},"David (Stanford 3D Scanning Repository)",{"dataset":103,"sequence":97,"environment":95},{"dataset":103,"sequence":99,"environment":95},{"dataset":103,"sequence":101,"environment":95},[108,112,114,116,118],{"name":109,"methodId":110,"linkable":111,"proposed":60,"self":60},"Poisson","kazhdan2006poisson",true,{"name":113,"methodId":5,"linkable":111,"proposed":111,"self":111},"new solver without screening (alpha = 0)",{"name":115,"methodId":54,"linkable":60,"proposed":60,"self":60},"Wavelet",{"name":117,"methodId":54,"linkable":60,"proposed":60,"self":60},"SSD",{"name":119,"methodId":5,"linkable":111,"proposed":111,"self":111},"Screened",[121,125,128,130,132,135,137,138,140,142,144,146,147,149,151,153,155,157,159,161,163,165,167,169,171,173,175,177,179,181,183,185,187,189,191,193,195,197,199,200,202,204,206,207,209,211,214,216,218,220,222,224,226,228,230,232,234,235,237,239,241,243,245,247,250,252,254,256,258,260,261],[122,122,122,123,124,122,122,124,122],0,10,-1,[126,126,122,127,124,122,122,124,122],1,13,[74,122,122,129,124,122,122,124,122],3,[129,122,122,131,124,122,122,124,122],275,[133,122,122,134,124,122,122,124,122],4,14,[122,74,122,136,124,122,122,124,122],113,[74,74,122,133,124,122,122,124,122],[129,74,122,139,124,122,122,124,122],238,[133,74,122,141,124,122,122,124,122],133,[122,122,126,143,124,122,122,124,122],25,[126,126,126,145,124,122,122,124,122],17,[74,122,126,133,124,122,122,124,122],[129,122,126,148,124,122,122,124,122],547,[133,122,126,150,124,122,122,124,122],20,[122,74,126,152,124,122,122,124,122],149,[74,74,126,154,124,122,122,124,122],11,[129,74,126,156,124,122,122,124,122],455,[133,74,126,158,124,122,122,124,122],269,[122,122,74,160,124,122,122,124,122],89,[126,126,74,162,124,122,122,124,122],36,[74,122,74,164,124,122,122,124,122],6,[129,122,74,166,124,122,122,124,122],3302,[133,122,74,168,124,122,122,124,122],44,[122,74,74,170,124,122,122,124,122],422,[74,74,74,172,124,122,122,124,122],35,[129,74,74,174,124,122,122,124,122],1247,[133,74,74,176,124,122,122,124,122],604,[122,122,129,178,124,122,122,124,122],320,[126,126,129,180,124,122,122,124,122],105,[74,122,129,182,124,122,122,124,122],9,[129,122,129,184,124,122,122,124,122],15441,[133,122,129,186,124,122,122,124,122],126,[122,74,129,188,124,122,122,124,122],1387,[74,74,129,190,124,122,122,124,122],118,[129,74,129,192,124,122,122,124,122],3495,[133,74,129,194,124,122,122,124,122],1622,[122,122,133,196,124,122,122,124,122],41,[126,126,133,198,124,122,122,124,122],45,[74,122,133,182,124,122,122,124,122],[129,122,133,201,124,122,122,124,122],492,[133,122,133,203,124,122,122,124,122],48,[122,74,133,205,124,122,122,124,122],427,[74,74,133,154,124,122,122,124,122],[129,74,133,208,124,122,122,124,122],863,[133,74,133,210,124,122,122,124,122],454,[122,122,212,213,124,122,122,124,122],5,108,[126,126,212,215,124,122,122,124,122],66,[74,122,212,217,124,122,122,124,122],12,[129,122,212,219,124,122,122,124,122],2355,[133,122,212,221,124,122,122,124,122],73,[122,74,212,223,124,122,122,124,122],510,[74,74,212,225,124,122,122,124,122],38,[129,74,212,227,124,122,122,124,122],1724,[133,74,212,229,124,122,122,124,122],932,[122,122,164,231,124,122,122,124,122],412,[126,126,164,233,124,122,122,124,122],157,[74,122,164,150,124,122,122,124,122],[129,122,164,236,124,122,122,124,122],19158,[133,122,164,238,124,122,122,124,122],182,[122,74,164,240,124,122,122,124,122],1498,[74,74,164,242,124,122,122,124,122],151,[129,74,164,244,124,122,122,124,122],4895,[133,74,164,246,124,122,122,124,122],2194,[122,122,248,249,124,122,122,124,122],7,1710,[126,126,248,251,124,122,122,124,122],522,[74,122,248,253,124,122,122,124,122],43,[133,122,248,255,124,122,122,124,122],609,[122,74,248,257,124,122,122,124,122],5318,[74,74,248,259,124,122,122,124,122],545,[129,74,248,54,122,122,122,124,122],[133,74,248,262,124,122,122,124,122],6188,[264],">8192 MB; exceeded available RAM",[79],[267],"laptop, quad-core Intel Core i7, 8 GB RAM",[],[270],"Wall-clock time and memory for Neptune and David at depths 8 to 11; screening weight alpha = 4, Neumann boundaries, samples-per-node 1; bracketed values are the new solver with alpha = 0; dagger: SSD at David depth 11 reports CPU user time because memory exceeded RAM",{"slug":272,"group":273,"sourceId":5,"sourceLabel":6,"table":274,"selfRows":133,"metrics":275,"seqs":281,"entrants":285,"cells":290,"outcomes":299,"locators":300,"hardware":304,"wordings":305,"notes":306},"kazhdan2013screened-text-sec-1","kazhdan2013screened:Text Sec. 1","Text Sec. 1",[276,278],{"label":277,"unit":84,"statistic":85,"alignment":85},"processing time without parallelization",{"label":279,"unit":280,"statistic":85,"alignment":85},"output triangles","million triangles",[282],{"dataset":283,"sequence":284,"environment":95},"David (Digital Michelangelo, 11.4M-point subset)","David head, depth 10",[286,288],{"name":287,"methodId":5,"linkable":111,"proposed":111,"self":111},"traditional Poisson (new implementation, screening weight 0)",{"name":289,"methodId":5,"linkable":111,"proposed":111,"self":111},"screened Poisson",[291,293,295,297],[122,122,122,292,124,122,124,124,122],230,[126,122,122,294,124,126,124,124,122],272,[122,126,122,296,124,74,124,124,122],6.8,[126,126,122,298,124,74,124,124,122],6.9,[],[301,302,303],"Sec. 1, footnote 1, Fig. 1","Sec. 1, Fig. 1","Sec. 1",[],[],[307],"Subset of 11.4M points from the David scan, octree depth 10 (effective 1024^3), timings without parallelization; unscreened timing uses the new implementation with screening weight 0",[],1790510665758]