【作者】 白云娜；

【导师】 蒋强荣； 张鸿宾；

【作者基本信息】 北京工业大学 ， 计算机科学与技术， 2011， 硕士

【摘要】 随着三维模型的造型、重构和可视化等相关技术的快速发展,以及先进的、高精度的扫描技术的不断推出,三维模型的复杂度和数据量急剧增长,远远超出了当前计算机实时的图形处理能力。因此产生了许多以模型分割为处理初步阶段的算法,以便大幅降低三维模型处理的复杂程度。另外,三维形状分析、模拟压缩、模型检索和三维动画等方面的应用上,三维模型的语义分割具有非常重要的实际意义。近年来,三维模型的分割引起了众多研究和实际工作者的关注,成为一个热点研究课题。本论文研究以视觉认知心理学为基础准则,通过对类人形三维模型的骨骼提取,依据人体测量学的先验知识完成三维骨骼的语义特征点提出,从而完成类人形三维模型的语义分割。本文的创新工作主要有:(1)详细研究了三维模型语义分割算法中的极小值规则和捷径规则,通过大量的实例分析了它的优点和局限性。(2)提出了一种基于骨架提取和特征点的三维模型语义分割算法。该算法采用将距离变换及势场方法相结合的一种线形骨架提取算法,提取模型骨骼线;利用人体测量学的先验知识为指导进行特征点的识别选取,从而使保证了语义分割的正确性;以视觉认知心理学为指导,利用骨骼线上的语义特征点完成对三维模型的语义分割。和3D捷径规则指导下的三维模型网格分割算法相比,本文算法的计算效率有了大幅提升。更多还原

【Abstract】 In last decades, 3D mesh reconstruction and visualization technology has an increasing development. With more advanced an accurate technology the equipment is employed, larger scale the 3d models have. More difficult of real-time processing is increasing with the enlarged data and complexity of model. Semantic segmentation is a curial process in computer animation. Many researches has been done, such as parameter, compression, morphing, which mostly using model segmentation algorithm as a preliminary stage. For mesh segmentation can greatly reduce the complexity of the algorithms.3D model segmentation became hot research topic in recent years. This paper presents a semantic segmentation algorithm under the guidance of skeleton of 3D humanoid model. The algorithm is in two phase: (1) extract the skeleton of humanoid model; (2) identify the feature points of the extracted skeleton and segment the model according the key point employ the short-cut rule of cross section.The main contributions of our work are:1) Research on the minima rule and shortcut rule, this paper analyzes the advantages and limitations of the minima rule and shortcut rule in detail through a lot of specific examples.2) This paper presents a semantic segmentation algorithm under the guidance of skeleton of 3D humanoid model. This paper extraction the skeleton line of the model using a curve-skeleton extraction algorithm combining distance transform and potential field, employ pre-knowledge of anthropometry to identification the semantic feature points of skeleton line which assures the correctness of semantic segmentation, finish the semantic segmentation of 3D model using the feature points based on visual cognitive psychology. Comparing with the semantic segmentation algorithm of 3d model in Li’s paper, the calculation efficiency of this algorithm is increased greatly.更多还原