shape modelling
play

Shape Modelling Aquisition Reconstruction Processing 17-06-2009 - PDF document

Shape data modelling and analysis as a support in the diagnosis of vascular diseases M. Attene, M. Mortara, G. Patan CNR-IMATI Ge Speaker: Michela Mortara michela@ge.imati.cnr.it Shape Modelling Aquisition Reconstruction


  1. Shape data modelling and analysis as a support in the diagnosis of vascular diseases M. Attene, M. Mortara, G. Patanè CNR-IMATI Ge Speaker: Michela Mortara michela@ge.imati.cnr.it Shape Modelling  Aquisition  Reconstruction  Processing  … 17-06-2009 Workshop on Anatomical Models, INRIA 2 1

  2. Shape Analysis  Characterization  Segmentation  Structuring  … 17-06-2009 Workshop on Anatomical Models, INRIA 3 In the following…  Tailor  Plumber with possible applications in the medical field  Convex Approximation  ShapeAnnotator 17-06-2009 Workshop on Anatomical Models, INRIA 4 2

  3. Tailor  Use a curvature analysis to compute the main features of a given shape and organize them into an abstract representation  Automate processes like matching, retrieval, comparison of shapes. M.Mortara, G.Patané, M.Spagnuolo, B.Falcidieno, J.Rossignac " Blowing Bubbles for the multiscale 17-06-2009 Workshop on Anatomical Models, INRIA 5 decomposition of triangle meshes " Algorithmica - special issue on shape algorithms. Vol 38, (1), pp. 227-248, Springer, 2003. Tailor  Label of v at scale i ← configuration of the intersection between the mesh and a sphere having Topology radius R i and centred in v. Curvature Geometric attributes 17-06-2009 Workshop on Anatomical Models, INRIA 6 3

  4. Classification criteria  Number of connected components:  1,2, 3 or more 17-06-2009 Workshop on Anatomical Models, INRIA 7 Classification criteria  One intersection curve  Curvature characterization:  Sharp  Rounded  Blend 17-06-2009 Workshop on Anatomical Models, INRIA 8 4

  5. Multi-scale curvature on meshes l : α = 2 π R : 2 π P α = l / R α Σ α i = L / R R  GC(v)=L/R ∈ [0, + ∞ ) l R  On a plane: GC(v)=2 π  On a spike: GC(v)< 2 π → 0  On a blend: GC(v)>2 π → + ∞  No distinction between convex and concave 17-06-2009 Workshop on Anatomical Models, INRIA 9 Classification criteria  Two intersection curves  Relative length characterization:  Cylindrical L max ≤ 2 L min  Conic otherwise 17-06-2009 Workshop on Anatomical Models, INRIA 10 5

  6. Classification criteria  Status:  One int.  Convex  Concave  More int.  Full  Empty 17-06-2009 Workshop on Anatomical Models, INRIA 11 Final classification Tip Tip Tip Tip Cylindrical Cylindrical Saddle Saddle Cylindrical Cylindrical Conic Conic Branching Branching TIP MOUNT PIT DIP Concave Concave BLEND LIMB JOINT FUNNEL Smooth Smooth WELL SPLIT HOLLOW 17-06-2009 Workshop on Anatomical Models, INRIA 12 6

  7. Query Language  Each vertex is described by a vector of labels  The i th label corresponds to the vertex characterization at scale R i 17-06-2009 Workshop on Anatomical Models, INRIA 13 Tailor results  Multi-scale Morphological analysis of the cortex 17-06-2009 Workshop on Anatomical Models, INRIA 15 7

  8. Curvature-based Skeleton 17-06-2009 Workshop on Anatomical Models, INRIA 16 ProTailor  Parallel implementation on a Linux Beowulf Cluster of 16 PCs  2.66 GHz Pentium IV processor  1 GB of Ram  2 EIDE 80 GB disks interfaced in RAID 0  by Antonella Galizia,IMATI  The scalability of the algorithm is almost linear (molecule model, 35MB, 31 minutes -> 2 minutes) M.Mortara, A.Galizia “ProTailor: a paralle operator for extremely fast shape analysis in Bioinformatics applications” 17-06-2009 Workshop on Anatomical Models, INRIA 17 in Proceedings of PDP2009, special session on Bioinformatics applications. 8

  9. Tailor on volumes  on tetrahedral meshes, computes the sphere surface inside the volume 17-06-2009 Workshop on Anatomical Models, INRIA 18 Plumber  Segmentation into tubular features and “bodies”  Is based on the Tailor characterization  Works in a multi-scale fashion wrt tube section size  Computes axis and sections of each tubular feature 17-06-2009 Workshop on Anatomical Models, INRIA 19 9

  10. Plumber The algorithm works in the following steps:  Selection of the scale R  Classification of vertices and identification of seed limb region  Tubular feature extraction  Increase R and repeat M. Mortara, G. Patané, M. Spagnuolo, B. Falcidieno, and J. Rossignac . Plumber: A Multi- 17-06-2009 Workshop on Anatomical Models, INRIA 20 scale Decomposition of 3D Shapes into Tubular Primitives and Bodies , Proc. of Solid Modeling and Applications, 2004 Plumber  A sphere is centred in the barycentre of the medial loop.  The sphere rolls in the two possible directions until a halting condition occurs. 17-06-2009 Workshop on Anatomical Models, INRIA 21 10

  11. Plumber  At each step, the sphere generates a new cross section and a new axis point.  Stop when:  The tube ends  Curve length over threshold  Bifurcation 17-06-2009 Workshop on Anatomical Models, INRIA 22 Results 17-06-2009 Workshop on Anatomical Models, INRIA 23 11

  12. Results 17-06-2009 Workshop on Anatomical Models, INRIA 24 Plumber on Point Clouds 17-06-2009 Workshop on Anatomical Models, INRIA 25 12

  13. Identification of human body parts 17-06-2009 Workshop on Anatomical Models, INRIA 26 Fitting Primitives  Generalization of the hierarchical face clustering (Garland et al. 01) Automatic generation of a binary tree of clusters,  each fitted to one of the available primitives: planes, spheres and cylinders  At the first step, each triangle is considered to be a cluster  Then, adjacent triangles are clustered according to cluster-to-primitive fitting 17-06-2009 Workshop on Anatomical Models, INRIA 27 M. Attene, B. Falcidieno, and M. Spagnuolo . Hierarchical Mesh Segmentation based on Fitting Primitives . The Visual Computer, 22, 2006 13

  14. Hierarchical Convex Approximation  Segments a shape into a hierarchy of nearly convex parts  Works on tetrahedra  Applications:  fast region selection from complex models  shape segmentation  shape approximation  deformation, editing Marco Attene, Michela Mortara, Michela Spagnuolo and Bianca Falcidieno “Hierarchical Convex Approximation 17-06-2009 Workshop on Anatomical Models, INRIA 28 of 3D Shapes for Fast Region Selection” Computer Graphics Forum, Vol. 27, No. 5 (SGP'08 Procs.), pp. 1323- 1333, 2008 Region selection  The tree of clusters can be traversed 17-06-2009 Workshop on Anatomical Models, INRIA 30 14

  15. Complex Selections  Tracking cutting lines or surface strokes can be complicated  Difficult topology  Occlusions Here the heart is connected with multiple vessels and tissues, and is tightly occluded by the chest, thus selecting it through cutting lines would be rather unpractical. Through our mechanism the selection required just a mouse click and a wheel rotation. 17-06-2009 Workshop on Anatomical Models, INRIA 31 Annotation  Psychological surveys show that humans “understand” shapes by recognizing interesting sub-parts and their structure (Marr 1982, Biederman 1987). Shape Segmentation   For specific contexts, it is possible to describe what these “features” are and how they are structured.  Geometric Description of the features  Structural Description of the shape  Semantic Annotation cylinder Cylinder Through IN plane Hole plane 17-06-2009 Workshop on Anatomical Models, INRIA 35 15

  16. ShapeAnnotator  User knowledge → usable explicit content Shape Abstracted A surface mesh Shape Shape CG Tools A segmented mesh Annotator Segmentation Plug-ins Domain An OWL ontology Instance Instance Expert Expert Knowledge Base 17-06-2009 Workshop on Anatomical Models, INRIA 36 ShapeAnnotator  The ShapeAnnotator is an open-source software project hosted by sourceforge.net  http://shapeannotator.sourceforge.net 17-06-2009 Workshop on Anatomical Models, INRIA 37 16

  17. Mesh Scenario simplification Part selection Aquisition/reconstruction Annotation Analysis … … Search/retrieval … … Comparison … Sharing … 17-06-2009 Workshop on Anatomical Models, INRIA 38 Applications  Monitoring shape changes over time  Comparison of the shape of anatomical regions among patients  Semantic rendering of anatomical regions and surgery planning  Analysis, automatic extraction of metadata, automatic or supported annotation of anatomical regions for future retrieval, comparison and analysis of collected data. 17-06-2009 Workshop on Anatomical Models, INRIA 39 17

Recommend


More recommend