颅面五官分割与形态统计中的关键技术研究

项目名称： 颅面五官分割与形态统计中的关键技术研究

项目编号： No.61305032

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 刘晓宁

作者单位： 西北大学

项目金额： 24万元

中文摘要： 颅面五官形态主要研究面貌与颅骨五官的生长变化规律、外观形态规律及相互关系，是人类体貌规律研究的重要分支。五官形态是颅骨复原中的关键技术和难点，现有的形态数据大多基于X线片获取，统计范围小，且未涉及三维形态。随着人类生活水平的提高，颅面五官形态发生了不小的变化，急需更新的、更科学的五官形态数据。 本项目以颅面网格数据为基础，主要研究内容包括：①基于颅面边缘点的CPD（Coherent Point Drift）三维颅面数据自动配准技术；②基于SVDD的颅面五官分割技术；③采用多重约束的层次点对应方法进行五官各分区数据的规格化处理，实现不同颅面之间数据稠密度一致性；④建立基于统计的五官形态统计模型，发现不同人颅骨之间的本质差异点，分析蒙古人种不同年龄、不同性别的颅面五官形态规律。 本项目建立在多年积累的头颅图像集基础之上, 研究成果可为颅骨复原、医学整容等相关领域提供科学依据与应用基础

中文关键词： 颅面复原；颅骨形态统计；五官分区；颅骨性别鉴定；颅面数据规格化

英文摘要： The research of morphology of craniofacial features is mainly to study the growth law of face and skull, the law of facial appearance. It is an important branch of human body studying. The morphology of craniofacial features is the key technology, also difficult, of the facial reconstruction. The present morphology data are mostly based on the X-ray, so the statistical range is relatively small, and is not related to the three-dimensional morphology. With the improvement of people's living condition, many changes are happened on the morphology of craniofacial features, we need the newest and more scientific features of morphological data. Based on the craniofacial grid data, the main research contents in this project include: ①the dynamical registration of craniofacial based on CPD; ②the segmentation of craniofacial features based on SVDD; ③To realize different craniofacial between data consistency consistency, we use hierarchical multiple constraints corresponding method of normalization features each partition data; ④the foundation of features morphology statistical model. Based on the model, we can find the key differences between different skulls, and analyze the craniofacial morphological features of Mongoloid in different age and sex. This project is based on the data we have got by several years. The rese

英文关键词： craniofacial reconstruction；Skull morphology statistics；Facial features partition；skull sex determination；Craniofacial data normalization