特征点自动标定的颅面复原及其评估方法  

Landmarks Automatic Calibration Based Craniofacial Reconstruction and Results Assessment Method

作　　者：梁荣华[1] 叶钱炜[1] 古辉[1] 马祥音[1] 缪永伟[1] 

机构地区：[1]浙江工业大学计算机科学与技术学院,杭州310023

出　　处：《计算机辅助设计与图形学学报》2013年第3期322-330,共9页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(61070114;61202205;61272309);浙江省自然科学基金重点项目(Z1090630)

摘　　要：为了采用计算机技术从人头颅骨样本重构出个性化的三维表面面部特征,提出一种特征点自动标定的颅面复原算法.首先利用特征点自动标定算法对待复原颅骨模型进行特征点标定;然后在基于模板变形的颅面复原算法基础上引入Laplace坐标网格变形技术,以实现待复原颅骨模型的颅面复原.为了验证文中算法进行颅面复原结果的可靠性,提出一种采用计算颅面整体特征和局部特征相似度的颅面复原结果评估方法.实验结果表明,文中提出的特征点自动标定算法大大提高了颅面复原前期工作的效率,颅面复原算法对待复原颅骨的颅面复原取得了良好的复原效果;同时,颅面复原结果评估方法达到了主观与客观评价上的一致,对复原结果验证有良好的指导作用.3D craniofacial model reconstruction from skull data has been a challenging research topic for many years. In this paper, a new craniofacial reconstruction algorithm based on automatic calibration of feature points is presented. First, the feature points are calibrated on skull model by landmarks＇ automatic calibration algorithm. Then the Laplace coordinate based mesh deformation is introduced to the template based craniofacial reconstruction technique to achieve the craniofacial reconstruction of skull. In order to verify the reliability of the reconstruction results, an assessment method is presented, which combines the similarity of overall facial features and local facial features. According to the experimental results, the algorithm greatly improves the efficiency of the preliminary work of craniofacial reconstruction, and achieves a satisfactory result, meanwhile the assessment method for the reconstruction results obtains consensus on the subjective and objective evaluation, and is a good indicator for craniofacial reconstruction results.

关 键 词：颅面复原 颅面特征点 迭代最近点匹配 Laplace坐标 颅面相似度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]