Mixed-Integer Quadrangulation精选.pptVIP

Mixed-Integer Quadrangulation David Bomme Henrik Zimmer LeifKobbelt SIGGRAPH 09 1.Introduction 2.Previous Work Quality of Quadrangulation Individual element quality: rectangle. Orientation: The edges capture curvature in a natural way. Alignment: Sharp features. Global structure: Singularities: Vertices whose valence is not 4. Allieze P. Anisotropic polygonal remeshing. SIGGRAPH 03 Dong S. Spectral surface quadrangulation. SIGGRAPH 06 3.Contribution We propose an adaptive greedy solver for mixed integer problems. The globally smooth parametrization technique allows us to generate seamless quad meshes while satisfying various constraints like orientation, alignment. 4.Procedure Cross field N-Symmetry direction fields (Ray et al.N-Symmetry direction fields design.2008) Definition: an angle-field: a period-jump field: Energy: :The angle of the neighboring coordinate frame The cross field index of vertex: Only singularities of the cross field have nonzero index which is always a multiple of ?.[Ray et al.2008] 1/4: valence 3 -1/4: valence 5 Finding a smooth, interpolating cross field: Minimizing the energy with feature constrains [Hildebrant et al.2005]: A greedy Mixed-Integer Solver Iteratively solve the linear problem and set Global Parametrization from smooth cross field Connect the integer node in the uv domain -Quad Mesh (Unproject) Piecewise linear scalar fields u and v defined on mesh. Define energy: and are two orthogonal vectors in T corresponding to the cross field directions Cut edge due to singularities Minimize the energy ,while the u v scalar value on singularities should be integer. It’s also a mixed-integer problem, which can be solved by the greedy mixed-integer solver. 5.Result Webpage: