user-guided simplification

Download User-Guided Simplification

Post on 31-Dec-2015




1 download

Embed Size (px)


User-Guided Simplification. Youngihn Kho Michael Garland University of Illinois at Urbana–Champaign. Surface Simplification. 5,804 faces. 780 faces. Previous simplifications. Most of previous algorithms are fully automatic generally work well, but weak at higher-level semantics - PowerPoint PPT Presentation


  • User-Guided Simplification

    Youngihn KhoMichael GarlandUniversity of Illinois at UrbanaChampaign

  • Surface Simplification5,804 faces780 faces

  • Previous simplificationsMost of previous algorithms are fully automaticgenerally work well, butweak at higher-level semanticsuser-guidance can improve the quality

    Semi-automatic methodsZeta (Cignoni et al. 97)Semisimp (Li and Watson 01)User-Controlled Creation of Multiresolution Meshes (Pojar and Schmalstieg 03)

  • User Guidance can be UsefulWe propose a user-guided simplificationdirectly guide simplification processesusers freely interact at any level

    Built on top of quadric-based simplification (Garland and Heckbert 97)

  • Iterative Edge Contraction Starting from the original model, iterativelyrank all edges with some cost metriccontract minimum cost edgeupdate edge costsgenerate vertex tree

  • Quadric-Error MetricGiven a plane, a quadric Q is defined asSquared distance of point to plane isSum of quadrics represents set of planes

  • Quadric-Error MetricEach edge has an associated quadric sum of quadrics for its two verticesfind a vertex v* minimizing Q(v*)After the edge contraction the vertex v* accumulates the associated quadrics

  • How to Guide Simplification:Manipulate QuadricsIn the quadric-based algorithmcontraction order and optimal positions are crucialquadrics determine both

    We manipulate quadrics in two main waysweighting quadricsadding constraint quadrics

  • Weighting Quadrics:Control Contraction CostsWe weight quadricsHeavy weights are appliedon feature areasincrease contraction costs

    Also changes optimal positionsa desirable side effectFeature areas are painted

  • Constraint Quadrics:Control Optimal PositionsWe can define constraint planes add their quadrics to appropriate vertices bias optimal positionsincrease contraction costs -> store separately

  • We Propose Three types ofConstraint QuadricsContour ConstraintsPoint ConstraintsPlane Constraints

  • Example : Contour Constraint

  • Example : Point Constraint

  • We Need:New Propagation RulesUsers want to freely interact at any levelaffect both simplification and refinement processvertex tree structures may be changedconstraint quadrics and representative weights should be appropriately propagated.

  • Propagation:Constraint Quadrics

  • Propagation:Representative Weight

  • Error Analysis:Relative Error DistributionFully automaticUser-guided

  • ConclusionNew interactive simplification systemextends an existing QSlim algorithmallows user-guidance to improve approximationslittle user effort, still efficientChanges vertex tree structurescan be used for further applications

  • Future WorkBetter weights selectioncurrently chosen by usersautomatic suggestion would be goodPerceptually based error metriclow-level perception models (Reddy 97, Luebke and Hallen 01)but also high-level : eg. actual feature vs noise machine learning techniques?

  • The EndThanks!

    Contact Information Youngihn kho Michael Garland