Sketch-Based 3D Shape Creation And Analysis

Digital Geometry is now widely applied to industry, entertainment and education business, so research on creating and analysis of geometric content become more and more popular. There are various ways to acquire and create digital geometric models,which mainly lie in two categories: one is using a depth sensor to scan the real object and reconstructing the surface approximating the scanned point cloud; the other is buildfrom-scratch, that it is, creating 3D conceptual models by interactive tools. The later one always appears in CAD design, game/animation character and scene creation. It basically starts from sketching of a set of 3D parametric curves and ends with 3D curve network,following by a surfacing step.Many digital geometry application, including shape correspondence, parametrization, texture mapping, remeshing/quad-meshing, depend on geometry analysis(most of time, geometry segmentation). A lot of researchers define segmentation from various perspectives, however, semantically meaningful geometry segmentation still remains as a hard problem. From geometric view, a semantical segmentation means, it is homogeneous everywhere within each segmented patch and also the boundaries between patches are aligned with silent features of the shape. In this paper, we propose two different segmentation algorithm according to user’s intended, one is automatic, and the other is user-interactive.The main contribution of our paper includes:1: Generating surfaces from 3D curve networks has been a longstanding problem in computer graphics. In this work we present a new algorithm for finding cycles that bound surface patches. Unlike prior art in this area, the output of our technique is unrestricted,generating both manifold and non-manifold geometry with arbitrary genus. The novel insight behind our method is to formulate our problem as finding local mappings at the vertices and curves of our network, where each mapping describes how incident curves are grouped into cycles. This approach lends us the efficiency necessary to present our system in an interactive design modeler, whereby the user can adjust patch constraints and change the manifold properties of curves while the system automatically re-optimizes the solution.2: Cutting along the feature band of a mesh creates more semantically meaning segmentation. Unlike spectral clustering, which requires input of number of clusters, we derive a non-parametric partition algorithm based on correlation clustering. To achieve this, we define graph with both positive and negative weights. The positive weight is the length of the dual of mesh edge computed by an anisotropic metric. The negative weight happens when it is a feature edge. We define ridge and valley line as features, and their strength as negative weight. Then we design a new iterated min-cut algorithm to find the minimal weighted cut. The algorithm is faster and more accurate than state-of-art correlation algorithms.3: Interactive Mesh Segmentation methods always balance between accuracy and complexity. Most of methods either allow fast segmentation, however lack of controlling of boundary, or are very time consuming when the result are accurate. We present an interactive method for mesh segmentation that is inspired by the classical live-wire interaction for image segmentation. The core contribution of the work is the definition and computation of wires on surfaces that are likely to lie at segment boundaries. We define wires as geodesics in a new tensor-based anisotropic metric, which improves upon previous metrics in stability and feature-awareness. We further introduce a simple but effective mesh embedding approach that allows geodesic paths in an anisotropic path to be computed efficiently using existing algorithms designed for Euclidean geodesics. Our tool is particularly suited for delineating segmentation boundaries that are aligned with features or curvature directions, and we demonstrate its use in creating artist-guided segmentations.