-
摘要: 在基于模板变形的颅面复原方法中，复原的开始阶段需要在数据库中选取与待复原颅骨最为相似的参考颅骨.鉴于基于三维模型的检索算法时间久且颅骨间的差异细微，从而不同于一般三维模型数据库中各模型的差异.因此，已有的三维模型检索算法不适用于颅骨检索.本文提出一种夹角信息和距离信息融合的颅骨轮廓特征提取算法，并在此基础上提出一种能够反映颅骨空域信息的剖面特征提取算法.检索时首先获取三维颅骨的正交投影和深度投影，通过正交投影获取轮廓的角度和距离特征，通过深度投影获得具有空域信息的剖面特征；然后对多个特征进行加权融合搜索到最相似颅骨；最后通过ICP+TPS对检索到的颅骨进行误差评估.实验表明，本算法在保证检索效率的同时，可以准确地应用在颅面复原前期对最相似颅骨的选择上.Abstract: In the template-based method of craniofacial reconstruction, the first step is to select the reference skull in the database which is the most similar to the skull being restored. 3D model retrieval algorithms usually take a long time, and the differences between skulls are so tiny that the general models in 3D model database cannot differentiate them, so any general 3D retrieval algorithm is not suitable for 3D skull retrieving. To cope with these problems, an algorithm for skull contour feature extraction with integration of angle and distance information is proposed in this paper. On the basis of this, a new profile feature extraction algorithm reflecting the spatial information of skull is also proposed. In the process of retrieving the most similar skull, firstly, orthogonal projection and depth projection of 3D skull are obtained. The angle and distance features of profiles are obtained by orthogonal projection while the profile features of spatial information are obtained by depth projection. Then, multiple features are fused according to their weights to search for the most similar skull. Finally, the error of the retrieved skull is evaluated by ICP + TPS algorithm. Experimental results show that the algorithm can ensure retrieval efficiency. Meanwhile, it can be accurately applied to craniofacial reconstruction of earlier stage to find out the most similar skull.
-
Key words:
- Skull similarity /
- craniofacial restoration /
- 2D profiles /
- spatial feature /
- multiscale projection
1) 本文责任编委 潘泉 -
图 6 $X$和$Y$两颅骨对应特征的加权差异
Fig. 6 The weighted differences of the corresponding features for $X$ skull and $Y$ skull
图 7 模板颅骨Skull_112和检索出来的最相似的四个颅骨
Fig. 7 Template skull_112 and four most similar retrieval skull
图 8 本文方法与Skull_112依次最相似的颅骨配准效果
Fig. 8 The successively most similar skull registration effect compared with Skull_112 using our method
图 9 各种方法检索出的与Skull_112依次最相似的4个颅骨
Fig. 9 Skull_112 and four most similar retrieval skull using various methods
图 10 各种方法与Skull_112颅骨依次最相似的颅骨配准效果
Fig. 10 The successively most similar skull registration effect compared with Skull_112 using various methods
表 1 本文方法与Skull_112颅骨最相似的四个颅骨的$Dist$
Table 1 The $Dist$ of four most similar skulls compared with Skull_112 using the method of this paper
参考颅骨编号 $Dist$ Skull_112 0 Skull_12 0.14 Skull_15 0.24 Skull_11 0.36 
下载: 导出CSV
-
[1] Liang R H, Lin Y L, Li J, Bao J R, Huang X P. Craniofacial model reconstruction from skull data based on feature points. In: Proceedings of the 11th IEEE International Conference on Computer-Aided Design and Computer Graphics. Huangshan, China: IEEE, 2009. 602-605 [2] Pei Y R, Zha H B, Yuan Z B. The craniofacial reconstruction from the local structural diversity of skulls. Computer Graphics Forum, 2008, 27 (7):1711-1718 doi: 10.1111/cgf.2008.27.issue-7 [3] 税午阳, 周明全, 武仲科, 邓擎琼.数据配准的颅骨面貌复原方法.计算机辅助设计与图形学学报, 2011, 23 (4):607-614 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjf201104007&dbname=CJFD&dbcode=CJFQShui Wu-Yang, Zhou Ming-Quan, Wu Zhong-Ke, Deng Qing-Qiong. An approach of craniofacial reconstruction based on registration. Journal of Computer-Aided Design and Computer Graphics, 2011, 23 (4):607-614 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjf201104007&dbname=CJFD&dbcode=CJFQ [4] Duan F Q, Huang D H, Tian Y, Wu Z K, Zhou M Q. 3D face reconstruction from skull by regression modeling in shape parameter spaces. Neurocomputing, 2015, 151:674-682 doi: 10.1016/j.neucom.2014.04.089 [5] 朱新懿, 耿国华.颅面重构中颅面相似度比较.计算机应用研究, 2010, 27 (8):3153-3155 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsyj201008096&dbname=CJFD&dbcode=CJFQZhu Xin-Yi, Geng Guo-Hua. Craniofacial similarity comparison in craniofacial reconstruction. Application Research of Computers, 2010, 27 (8):3153-3155 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsyj201008096&dbname=CJFD&dbcode=CJFQ [6] 王占松, 田凌.基于功能的三维模型检索系统.计算机辅助设计与图形学学报, 2013, 25 (12):1877-1885 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjf201312014&dbname=CJFD&dbcode=CJFQWang Zhan-Song, Tian Ling. Function-based 3D model retrieval system. Journal of Computer-Aided Design and Computer Graphics, 2013, 25 (12):1877-1885 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjf201312014&dbname=CJFD&dbcode=CJFQ [7] 周燕, 曾凡智, 杨跃武.基于多特征融合的三维模型检索算法.计算机科学, 2016, 43 (7):303-309 doi: 10.11896/j.issn.1002-137X.2016.07.056Zhou Yan, Zeng Fan-Zhi, Yang Yue-Wu. 3D model retrieval algorithm based on multi feature fusion. Computer Science, 2016, 43 (7):303-309 doi: 10.11896/j.issn.1002-137X.2016.07.056 [8] 胡晓彤, 王建东.基于子空间特征向量的三维点云相似性分析.红外与激光工程, 2014, 43 (4):1316-1321 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=hwyj201404053&dbname=CJFD&dbcode=CJFQHu Xiao-Tong, Wang Jian-Dong. Similarity analysis of three-dimensional point cloud based on eigenvector of subspace. Infrared and Laser Engineering, 2014, 43 (4):1316-1321 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=hwyj201404053&dbname=CJFD&dbcode=CJFQ [9] 李海生, 张朝立, 蔡强, 毛典辉, 杜军平.基于信息熵加权的三维模型特征融合算法.计算机研究与发展, 2014, 51(S):57-68 http://www.wanfangdata.com.cn/details/detail.do?_type=conference&id=8863941Li Hai-Sheng, Zhang Chao-Li, Cai Qiang, Mao Dian-Hui, Du Jun-Ping. A 3D model feature fusion algorithm based on entropy weights. Journal of Computer Research and Development, 2014, 51 (S):57-68 http://www.wanfangdata.com.cn/details/detail.do?_type=conference&id=8863941 [10] Nie W Z, Liu A A, Su Y T. 3D object retrieval based on sparse coding in weak supervision. Journal of Visual Communication and Image Representation, 2016, 37:40-45 doi: 10.1016/j.jvcir.2015.06.011 [11] 周继来, 周明全, 耿国华, 王小凤.基于曲度特征的三维模型检索算法.计算机应用, 2016, 36 (7):1914-1917 doi: 10.11772/j.issn.1001-9081.2016.07.1914Zhou Ji-Lai, Zhou Ming-Quan, Geng Guo-Hua, Wang Xiao-Feng. 3D model retrieval algorithm based on curvedness feature. Journal of Computer Applications, 2016, 36 (7):1914-1917 doi: 10.11772/j.issn.1001-9081.2016.07.1914 [12] 杜卓明.基于改进的SKPCA三维模型检索方法.智能计算机与应用, 2014, 4 (5):29-31 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=dlxz201405009&dbname=CJFD&dbcode=CJFQDu Zhuo-Ming. 3D model retrieval based on the improved SKPCA. Intelligent Computer and Applications, 2014, 4 (5):29-31 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=dlxz201405009&dbname=CJFD&dbcode=CJFQ [13] Chen Z Y, Lin W C, Tsai C F, Ke S W. 3D model retrieval by sample based alignment. Journal of Visual Communication and Image Representation, 2016, 40:721-731 doi: 10.1016/j.jvcir.2016.08.017 [14] 孙晓鹏, 王冠, 王璐, 魏小鹏. 3D点云形状特征的二维主流形描述.软件学报, 2015, 26 (3):699-709 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=rjxb201503017&dbname=CJFD&dbcode=CJFQSun Xiao-Peng, Wang Guan, Wang Lu, Wei Xiao-Peng. 3D point cloud shape feature descriptor using 2D principal manifold. Journal of Software, 2015, 26 (3):699-709 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=rjxb201503017&dbname=CJFD&dbcode=CJFQ [15] 樊亚春, 谭小慧, 周明全, 郑霞.基于局部多尺度的三维模型草图检索方法.计算机学报, 2016, 39 (68):1-19 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjx201711002&dbname=CJFD&dbcode=CJFQFan Ya-Chun, Tan Xiao-Hui, Zhou Ming-Quan, Zheng Xia. A scale invariant local descriptor for sketch based 3D model retrieval. Chinese Journal of Computers, 2016, 39 (68):1-19 http://kns.cnki.net/KCMS/detail/detail.aspx?filename=jsjx201711002&dbname=CJFD&dbcode=CJFQ [16] Hu R X, Jia W, Ling H B, Huang D S. Multiscale distance matrix for fast plant leaf recognition. IEEE Transactions on Image Processing, 2012, 21 (11):4667-4672 doi: 10.1109/TIP.2012.2207391 [17] 汤昊林, 杨扬, 杨昆, 罗毅, 张雅莹, 张芳瑜.基于混合特征的非刚性点阵配准算法.自动化学报, 2016, 42 (11):1732-1743 http://www.aas.net.cn/CN/abstract/abstract18962.shtmlTang Hao-Lin, Yang Yang, Yang Kun, Luo Yi, Zhang Ya-Ying, Zhang Fang-Yu. Non-rigid point set registration with mixed features. Acta Automatica Sinica, 2016, 42 (11):1732-1743 http://www.aas.net.cn/CN/abstract/abstract18962.shtml [18] 陆雪松, 涂圣贤, 张素.一种面向医学图像非刚性配准的多维特征度量方法.自动化学报, 2016, 42 (9):1413-1420 http://www.aas.net.cn/CN/abstract/abstract18929.shtmlLu Xue-Song, Tu Sheng-Xian, Zhang Su. A metric method using multidimensional features for nonrigid registration of medical images. Acta Automatica Sinica, 2016, 42 (9):1413-1420 http://www.aas.net.cn/CN/abstract/abstract18929.shtml [19] 闫自庚, 蒋建国, 郭丹.基于SURF特征和Delaunay三角网格的图像匹配.自动化学报, 2014, 40 (6):1216-1222 http://www.aas.net.cn/CN/abstract/abstract18392.shtmlYan Zi-Geng, Jiang Jian-Guo, Guo Dan. Image matching based on SURF feature and Delaunay triangular meshes. Acta Automatica Sinica, 2014, 40 (6):1216-1222 http://www.aas.net.cn/CN/abstract/abstract18392.shtml [20] Besl P J, McKay N D. A method for registration of 3-D shapes. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1992, 14 (2):239-256 doi: 10.1109/34.121791 [21] Bookstein F L. Principal warps:thin-plate splines and the decomposition of deformations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1989, 11 (6):567-585 doi: 10.1109/34.24792 -
点击查看大图
图(11) / 表(2)
计量
- 文章访问数: 1437
- HTML全文浏览量: 196
- PDF下载量: 543
- 被引次数: 0
