一种评估颅骨和颅面相似性的方法  

Similarity measurement method of skull and craniofacial data

作　　者：晋武侠[1] 李康[1] 耿国华[1] 刘伦椿[1] 

机构地区：[1]西北大学信息科学与技术学院,西安710127

出　　处：《计算机应用研究》2013年第10期3124-3127,3151,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(61172170)

摘　　要：针对评估颅面和颅骨轮廓相似性过程中的离散点对应、曲线拟合带来的误差以及人工干预问题,提出一种基于重采样和Fourier变换的离散轮廓像素点的相似度量方法。首先使用Canny算子和滑动窗口方法进行边缘检测和边界跟踪,得到待比较颅骨和颅面的边缘轮廓像素集合;然后对轮廓统一重采样,解决了两者离散点的对应问题;最后直接对数据进行规格化和离散傅里叶变换,利用傅里叶描述子进行相似性度量,避免了曲线拟合导致的误差和手工介入问题。实验表明,该算法提高了颅骨和颅面相似性度量的准确率,复杂度较低,可实现评价过程自动化,为颅像重合身份认证奠定了基础。Traditional measurement method of skull and craniofacial similarity has some disadvantages,such as discrete point correspondence, the errors caused by curve fitting and the problem of manual intervention and so on. In order to overcome those problems, this paper proposed the measure method based on resampling and DFT（ discrete Fourier transform） of contour pixel. This paper first used Canny operator and sliding window method for edge detection and boundary tracking. Secondly, it took resampling on a pair of contour extracted from the edge of mandibular of skull and craniofacial image. Then it employed normalizing and discrete Fourier transform, measured the similarity of this pair of Fourier descriptor. Finally, it got a measurement value of determining similarity. Experimental results show that the method achieves a precise effect, has lower algorithm complexity, realizes automatical evaluation process and lays a foundation for skull-photo superimposition of identity authentication.

关 键 词：重采样 离散Fourier变换 颅像重合 相关系数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]