Technical University of Munich Process monitoring of construction sites by photogrammetric point clouds and 4D building information models (BIM) [selected slides] Uwe Stilla , stilla@ tum.de Co-Authors: S. Tuttas, A. Braun, A. Borrmann 1 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Motivation Monitoring of building construction progress based on photogrammetric point clouds and 4d-building information models (BIM) ProgressTrack: funded by German Research Fundation r Operating sequences of building projects are extremly dynamic r difficult predictable boundary conditions (e.g. weather) ƒ strong dependency of single process steps ƒ Very often strong delays and budget overrun � acquisition of the actual construction progress is essential for r cost controlling ƒ organization of the further working schedule ƒ up to now: r manual by inspection of the site manager ƒ documentation by a handwritten construction diary ƒ � planed: automatic acquisition 2 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Motivation – Building Information Modeling (BIM) http://buildipedia.com/aec-pros/design-news/the-daily-life-of-building-information-modeling-bim 4 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
GIS world vs. BIM world GIS BIM Surface based Volume based Geometry CityGML IFC Description Five levels None Level of detail Building elements Semantics Boundary surfaces (Gröger et al., 2012) (Nagel et al., 2009) 6 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Monitoring of construction progress - Overview B I M Photogrammmetry (as-planned state) (as-built state) Data acquistion Schedule Image orientierung 3d geometry and Updating Dense matching object atributes Co-registrierung 3d point cloud State T i State T i Comparison Action OK 7 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Crane camera 17 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
18 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Terrestrial aquisition 20 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Camera positions in space building camera positions 21 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Different positions of aquisition 22 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Positions of acquisitions JUN-27 23 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Positions of acquisitions JUN-12 24 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Image orientation – corresponding points Ground Control Points Tie points (Features) STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Ground Control Points – points having known coordinates 26 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Tie points – corresponding image points Tie points shown in space STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Generation of the 3d point cloud Selection of appropriate stereo pairs Depth value for each pixel by Dense-Stereo-Matching Fusion of depth maps → Triangulation of 3d points STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
29 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
as-planned / as-build - MAY-15 31 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
as-planned / as-build - JUN-17 32 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
as-planned / as-build - SEP–25 33 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
36 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Typische Punktverteilung an Bauteilen: Punkte auf Bauteiloberfläche r a) bei guter Tiefengenauigkeit, b) bei schlechterer Tiefengenauigkeit, c) bei Verdeckung des unteren Bereiches; d) Bauteil mit Verdeckungen und Punkte vom Gerüst; e) Bauteil mit Verschalung; f) nicht vorhandenes Bauteil bei vorhandenem temporären Objekt 38 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Measure for verification Which construction parts are r verified by the points? Extraction of points around r the object parts. Rasterization of object r planes (10 cm) Check for every raster cell if r the points confirm the presence of the construction object. Measure which considers amount of points, Values for thresholds: r r accuracy ( σ ) and distance ( d ) of points to plane: p d = 2 cm “Point density” d i = 2 cm = > d d für d d 1 1 ∑ = ⋅ = σ d = 1 cm i i i min mit M d µ ⋅ σ = ≤ i für d d d d d i i min i min d min = 0.5 cm d i d i 39 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
All results are available in the BIM ! 43 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Rekonstruction of scaffolds from point clouds Assumtion: Satandard parts of a scaffold are known (e.g. length and diameter of the poles) 45 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
RPAS – Remotly Piloted Aerial System RPAS ASCTEC r Falcon ƒ Camera Sony NEX-7 r RGB r 6000 x 4000 pix r Wide angle lens r 47 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Mosaic 49 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Aufnahme: 2014-DEZ-12 50 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Aufnahme – Schrägsicht - JAN-16 53 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
55 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Cameras mounted on the crane jib 59 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Cameras 60 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
61 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Baustelle: TranslaTUM crane camera: JAN-14 62 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Sensorsystem – Krankamera 64 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Sensorsystem - Krankamera 65 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
66 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Coverage - Crane Usually all areas can be reached by the cranes. r The higher the building the more cameras are necessary. r A controlled crane movement is necessary to ensure full r coverage. Trolley and hoisting rope may be in the view. r 67 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
69 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Verwendete Gitterstrukturen Voxelgitter über den ganzen Beobachtungs- bereich für die Sichtbarkeitsanalyse Rastergröße Gerasterte Bauteilfläche Kantenlänge Begrenzung der Voxel Pufferbereich für die Punkt- extraktion Breite des Pufferbereichs hinten vorne 73 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Aussagen anhand von Sichtbarkeitsbedingungen Unterscheidung von Bauteilen r (1) sichtbar und vorhanden (2) sichtbar und vorhanden (3) definitiv nicht vorhanden – da (2) sichtbar (4) keine Aussage möglich – da nicht sichtbar Kamera � � Kamera 3 Kamera 2 3D-Punkte geplantes Bauteil frei belegt � � unbekannt 74 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Summary of acquisitions Construction site Photogrammmetry Photogrammmetry Photogrammmetry (as-built state) (as-built state) (as-built state) mobile, terrestrial mobile, airborne mobile, static view handheld UAV (RPAS) crane cameras 3d point cloud 77 STILLA et al. (2017-09-15) Photogrammetric Week, Stuttgart
Recommend
More recommend
