We present an algorithm for segmenting a mesh into patches whose boundaries are aligned with prominent ridge and valley lines of the shape. Our key insight is that this problem can be formulated as correlation cluster...We present an algorithm for segmenting a mesh into patches whose boundaries are aligned with prominent ridge and valley lines of the shape. Our key insight is that this problem can be formulated as correlation clustering(CC), a graph partitioning problem originating from the data mining community.The formulation lends two unique advantages to our method over existing segmentation methods. First,since CC is non-parametric, our method has few parameters to tune. Second, as CC is governed by edge weights in the graph, our method offers users direct and local control over the segmentation result. Our technical contributions include the construction of the weighted graph on which CC is defined, a strategy for rapidly computing CC on this graph, and an interactive tool for editing the segmentation. Our experiments show that our method produces qualitatively better segmentations than existing methods on a wide range of inputs.展开更多
Feature lines are fundamental shape descriptors and have been extensively applied to computer graphics, computer-aided design, image processing, and non-photorealistic renderingi This paper introduces a unified variat...Feature lines are fundamental shape descriptors and have been extensively applied to computer graphics, computer-aided design, image processing, and non-photorealistic renderingi This paper introduces a unified variational framework for detecting generic feature lines on polygonal meshes. The classic Mumford-Shah model is extended to surfaces. Using F-convergence method and discrete differential geometry, we discretize the proposed variational model to sequential coupled sparse linear systems. Through quadratic polyno- mials fitting, we develop a method for extracting valleys of functions defined on surfaces. Our approach provides flexible and intuitive control over the detecting procedure, and is easy to implement. Several measure functions are devised for different types of feature lines, and we apply our approach to various polygonal meshes ranging from synthetic to measured models. The experiments demonstrate both the effectiveness of our algorithms and the visual quality of results.展开更多
基金supported in part by a gift from Adobe System, Inc
文摘We present an algorithm for segmenting a mesh into patches whose boundaries are aligned with prominent ridge and valley lines of the shape. Our key insight is that this problem can be formulated as correlation clustering(CC), a graph partitioning problem originating from the data mining community.The formulation lends two unique advantages to our method over existing segmentation methods. First,since CC is non-parametric, our method has few parameters to tune. Second, as CC is governed by edge weights in the graph, our method offers users direct and local control over the segmentation result. Our technical contributions include the construction of the weighted graph on which CC is defined, a strategy for rapidly computing CC on this graph, and an interactive tool for editing the segmentation. Our experiments show that our method produces qualitatively better segmentations than existing methods on a wide range of inputs.
文摘Feature lines are fundamental shape descriptors and have been extensively applied to computer graphics, computer-aided design, image processing, and non-photorealistic renderingi This paper introduces a unified variational framework for detecting generic feature lines on polygonal meshes. The classic Mumford-Shah model is extended to surfaces. Using F-convergence method and discrete differential geometry, we discretize the proposed variational model to sequential coupled sparse linear systems. Through quadratic polyno- mials fitting, we develop a method for extracting valleys of functions defined on surfaces. Our approach provides flexible and intuitive control over the detecting procedure, and is easy to implement. Several measure functions are devised for different types of feature lines, and we apply our approach to various polygonal meshes ranging from synthetic to measured models. The experiments demonstrate both the effectiveness of our algorithms and the visual quality of results.
