定制化人像产品三维建模知识库技术研究及应用

发布时间：2018-04-29 04:28

本文选题：定制化人像产品 + 三维建模　；参考：《重庆大学》2013年硕士论文

【摘要】：制造业企业只有不断开发各种个性化定制产品，以满足人们日益增加的个性化需求，才能赢得生存与发展。当前，一类与人像有关的定制化产品正逐渐形成一个新兴产业，该产业的关键核心技术就是如何快速依据二维图像建立相似度高的三维模型。目前人像三维建模最重要的方法是基于图像的人像建模，已被广泛应用于影视、广告、游戏、人物动画等领域。然而，现有方法主要依靠图像的特征点进行标准模型的宏观变形，设计师再依据图像一笔一划地修改模型细节，建模效率较低。由于，各个人像产品具有一定共性和相同部位，因此，可以首先利用知识库技术来加速人像产品共性特征的建模，减少不同人像建模的重复工作量。本文就是围绕如何将知识库技术引入定制化人像产品开发而展开研究，以提高人像产品建模效率，为其产业化提供一种理论方法和技术支持。研究内容主要包括以下几个方面：首先，研究了人像产品三维建模知识的获取。通过分析知识的类型及其获取方式和获取步骤，根据人像产品三维建模知识的内容分析了该领域知识的特点及其分类，经过问题识别、概念化、形式化、实现及检验五个阶段最终实现人像产品三维建模知识的获取，，并给出了人像产品三维建模知识的部分规则。其次，研究了人像产品三维建模知识的表示。通过分析现有的知识表示方法——产生式、框架和面向对象等知识表示方法及其特点，并结合对人像产品三维建模知识内容和特点的分析，提出了适合该领域知识表示的混合知识表示法，即运用面向对象思想封装产生式、框架而形成的表示方法，最终实现人像产品三维建模知识的表示，并分别给出了人像产品三维建模知识中规则知识和实例知识的表示形式。然后，研究了人像产品三维建模知识库的功能实现。通过分析现有人像产品三维建模系统的不足，并结合本课题组已开发的人像产品三维建模支持系统，在此基础上，提出了一种基于标准模板与筛选算法的人像产品快速建模方法；将此方法作为知识库关键技术集成到人像产品三维建模支持系统中，论述了集成知识库技术的人像产品三维建模系统在人像产品开发过程中的应用。最后，将基于标准模板与筛选算法的人像产品快速建模方法集成到基于图像的三维人像产品建模设计系统中，并在定制化真人工艺品项目中得到应用，验证了该人像建模方法的有效性。案例结果表明，本文研究的方法和开发的建模系统具有良好的应用效果，不仅明显地提高了建模的效率，也在一定程度上提高了建模的相似度。
[Abstract]:In order to win survival and development, manufacturing enterprises must develop various individualized products to meet the increasing demand of individuation. At present, a class of customized products related to human images is gradually forming a new industry. The key technology of this industry is how to quickly build 3D models with high similarity based on two-dimensional images. At present, the most important method of 3D modeling of portrait is image-based portrait modeling, which has been widely used in film and television, advertising, games, character animation and other fields. However, the existing methods mainly rely on the feature points of the image to carry out the macroscopic deformation of the standard model. The designer modifies the details of the model one stroke at a time according to the image, so the modeling efficiency is relatively low. Because each portrait product has certain commonness and the same part, so we can use the knowledge base technology to accelerate the modeling of the common characteristics of the portrait product, and reduce the duplicate workload of the different portrait modeling. This paper focuses on how to introduce knowledge base technology into customized portrait product development in order to improve the efficiency of human image product modeling and provide a theoretical method and technical support for its industrialization. The research mainly includes the following aspects: Firstly, the acquisition of 3D modeling knowledge of portrait products is studied. By analyzing the types of knowledge, their acquisition methods and acquisition steps, according to the content of 3D modeling knowledge of portrait products, this paper analyzes the characteristics and classification of knowledge in this field, through problem identification, conceptualization, formalization, Finally, the acquisition of 3D modeling knowledge of humanoid products is realized in the five stages of realization and testing, and some rules of 3D modeling knowledge of humanoid products are given. Secondly, the representation of 3D modeling knowledge of portrait products is studied. Based on the analysis of the existing knowledge representation methods, such as production, framework and object oriented, and combining with the analysis of the content and characteristics of 3D modeling knowledge of portrait products, In this paper, a hybrid knowledge representation method suitable for knowledge representation in this field is proposed, that is, the representation method formed by using object oriented thought to encapsulate the production style and the framework, and finally realize the representation of 3D modeling knowledge of human portrait products. The representation forms of rule knowledge and case knowledge in 3D modeling knowledge of portrait products are given respectively. Then, the function realization of 3D modeling knowledge base of portrait product is studied. Based on the analysis of the deficiency of the existing 3D modeling system of portrait product and the 3D modeling support system of human portrait product which has been developed by our research group, This paper presents a rapid modeling method of human portrait product based on standard template and screening algorithm, which is integrated into 3D modeling support system of portrait product as a key technology of knowledge base. This paper discusses the application of 3D modeling system of human portrait product based on integrated knowledge base technology in the process of portrait product development. Finally, the rapid modeling method based on standard template and screening algorithm is integrated into the 3D human portrait product modeling and design system based on image, and applied in the customized human handicraft project. The validity of the method is verified. The results of the case show that the method and modeling system developed in this paper have good application effect, which not only obviously improve the efficiency of modeling, but also improve the similarity of modeling to a certain extent.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.41

【参考文献】