结合改进卷积神经网络和最小二乘法的颅骨性别鉴定

doi:10.16359/j.cnki.cn11-1963/q.2018.0030

人类学学报 ›› 2019, Vol. 38 ›› Issue (02): 265-275.doi: 10.16359/j.cnki.cn11-1963/q.2018.0030

结合改进卷积神经网络和最小二乘法的颅骨性别鉴定

杨稳(), 刘晓宁^*(), 刘雄乐, 朱丽品

西北大学信息科学与技术学院,西安 710127

收稿日期:2017-08-01 修回日期:2018-02-22 出版日期:2019-05-15 发布日期:2020-09-10
通讯作者: 刘晓宁
作者简介:杨稳（1993-）,男,陕西安康人,主要从事机器学习与模式识别研究,Email: 920759301@qq.com。
基金资助:
国家自然科学基金项目(61673319);国家自然科学基金项目(61363065);陕西省自然科学基金项目(2018JM6061);陕西省自然科学基金项目(2014JM8358);研究生自主创新项目(YZZ17181)

Skull sex identification using improved convolution neural network and least squares method

YANG Wen(), LIU Xiaoning^*(), LIU Xiongle, ZHU Lipin

College of Information Science and Technology, Northwest University, Xian, 710127

Received:2017-08-01 Revised:2018-02-22 Online:2019-05-15 Published:2020-09-10
Contact: LIU Xiaoning

摘要/Abstract

摘要：

颅骨性别鉴定在法医学和颅骨面貌复原等领域具有重要研究意义和应用价值,针对传统颅骨性别鉴定需要专家参与且主观性强、计算机辅助方法需要人工标定特征点等问题,本文提出了结合改进卷积神经网络和最小二乘法的颅骨性别鉴定方法。首先,获取三维颅骨模型多角度颅骨图像,利用改进的卷积神经网络计算每个样本的每张图像属于男性和女性的概率;其次,基于概率均值采用最小二乘法计算每张图像对性别鉴定的权重;最后,利用上述步骤得到的最优参数构造决策函数,通过决策值完成颅骨性别鉴定。本文方法抛弃了繁琐的手动测量,对完整颅骨的性别鉴定正确率高达94.4%,对不完整颅骨的性别鉴定正确率高达87.5%,能够获得较好的颅骨性别鉴定性能。

关键词: 颅骨性别鉴定, 多角度图像, 改进卷积神经网络, 最小二乘法

Abstract:

Skull sex identification has significant research and applied value in forensic anthropology and skull reconstruction. The traditional skull sex determination methods need expert participation and to some extent, is not objective, because computer-aided methods require to marking landmarks the feature points manually. We present a novel sex determination method based on improved Convolution Neural Network and Least Square. Firstly, obtain multi-angle skull images of three-dimensional skull model, and calculate the probability of each image belongs to male or female. Secondly, the weight of each image is calculated using the Least Squares method based on the probability mean. Lastly, the sex determination function is constructed by using the optimal parameters obtained through the above steps. This method does not need to mark the feature points or do the measurement. Experiments show that the proposed method can get quite a reliable performance with an accuracy of 94.4% for the the complete skull and 87.5% for the incomplete skull.

Key words: Sex determination, Imagery, improved convolution neural network, Least squares method

中图分类号:

Q989

杨稳, 刘晓宁, 刘雄乐, 朱丽品. 结合改进卷积神经网络和最小二乘法的颅骨性别鉴定[J]. 人类学学报, 2019, 38(02): 265-275.

YANG Wen, LIU Xiaoning, LIU Xiongle, ZHU Lipin. Skull sex identification using improved convolution neural network and least squares method[J]. Acta Anthropologica Sinica, 2019, 38(02): 265-275.

图/表 8

图1 本文的算法流程图

Fig.1 The algorithm flow chart of this paper

图2 归一化坐标图 a)法兰克福坐标系 /Frankfurt Coordinates ; (b)颅骨法兰克福坐标系Skull Frankfurt coordinate system

Fig.2 Normalized coordinate map

图3 改进LeNET5架构

Fig.3 Improved LeNET5 architecture

图4 学习速率为0.01时的分类结果

Fig.4 The learning rate was 0.01 for the classification results

图5 学习速率为0.001时的分类结果

Fig.5 The learning rate was 0.001 for the classification results

图6 学习速率为0.01时的分类结果

Fig.6 The learning rate was 0.01 for the classification results

图7 学习速率为0.001时的分类结果

Fig.7 The learning rate was 0.001 for the classification results

图8 不同学习速率下颅骨性别分类结果比较 a)学习速率为0.0005的分类效果/The learning rate was 0.0005 for the classification results; b)学习速率为0.001的分类效果/The learning rate was 0.001 for the classification results; c)学习速率为0.01的分类效果/The learning rate was 0.01 for the classification results; d)学习速率为0.05的分类效果/The learning rate was 0.05 for the classification results

Fig.8 Comparison of sex classification results of skull under different learning rates

参考文献 21

[1]	Guyomarc’H P, Bruzek J. Accuracy and reliability in sex determination from skulls: a comparison of Fordisc® 3.0 and the discriminant function analysis[J]. Forensic Science International, 2011, 208(1): 180.e1-180.e6
[2]	Stewart TD. Medico-legal aspects of the skeleton;sex,age,race and stature[J]. American Journal of Physical Anthropology, 1948,6(3):315-321 doi: 10.1002/ajpa.1330060306 URL pmid: 18884224
[3]	Iscan MY, Maryna S. The Human Skeleton in Forensic Medicine[M]. Royal Society of Medicine Press, 2013, 56(2): A-48
[4]	Keen JA. A study of the differences between male and female skull[J]. American Journal of Physical Anthropology, 1997,116(2):118-124
[5]	张亚盟, 魏偏偏, 吴秀杰. 现代人头骨断面轮廓的性别鉴定——基于几何形态测量的研究[J]. 人类学学报, 2016,35(2):172-180
[6]	Ogawa Y, Imaizumi K, Miyasaka S, et al. Discriminant functions for sex estimation ofmodern Japanese skulls[J]. Journal of Forensic & Legal Medicine, 2013,20(4):234-238 doi: 10.1016/j.jflm.2012.09.023 URL pmid: 23622466
[7]	Kanchan J, Gupta A, Krishan K. Estimation of sex from mastoid trsiangle-A craniometric analysis[J]. Journal of Forensic & Legal Medicine, 2013,20(7):855-860 doi: 10.1016/j.jflm.2013.06.016 URL pmid: 24112336
[8]	税午阳, 殷荣超, 周明全, 等. 中国汉族人颅骨数字模型的性别判别方法[J]. 中国法医学杂志, 2013,286(6):461-464
[9]	Twisha S, Patel MN, Nath S, et al. Determination of sex using cephalo-facial dimensions by discriminant function and logistic regression equations[J]. Egyptian Journal of Forensic Sciences, 2016,6(2):114-119 doi: 10.1016/j.ejfs.2015.03.004 URL
[10]	Santos F, Guyomarc’h P, Bruzek J. Comparison of linear discriminant analysis, logistic regression, support vector machines[J]. Forensic Sci Int, 2014, 245(204): el-e8
[11]	Alunni V, Jardin PD, Nogueira L, et al. Comparing discriminant analysis and neural network for the determination of sex using femur head measurements[J]. Forensic Sci Int, 2015,253:81-87 doi: 10.1016/j.forsciint.2015.05.023 URL pmid: 26093772
[12]	Afrianty I, Nasien D, Kadir MRA, et al. Back-Propagation neural network for gender determination in forensic anthropology[J]. Studies in Computational Intelligence, 2015,575:255-281
[13]	Luo L, Wang M, Tian Y, et al. Automatic Sex Determination of Skulls Based on a Statistical Shape Model[J], Computational and mathematical methods in medicine, 2013,2013(1):251628
[14]	Janarthanan R, Asha J. Supraorbital Margins for Identification of Sexual Dimorphism and Age Detection from Human Skull Using Wavelets[J]. Asian Journal of Applied Science and Technology, 2017,1(2):279-282
[15]	任荣荣, 周明全, 耿国华, 等. 三维颅骨形态量化表示与非线性性别判定[J]. 北京师范大学学报(自然科学版), 2017,53(1):19-23
[16]	Pham DV. Online handwriting recognition using multi-convolution neural network[J]. International Conference on Simulated Evolution & Learning, 2012,7673:310-319
[17]	Lawrence S, Giles CL, Tsoi AC, et al. Face recognition: a convolutional neural-network approach[J]. IEEE Transactions on Neural Network, 1997,8(1):98-113 doi: 10.1109/72.554195 URL
[18]	Hu B, Lu Z, Li H, et al. Convolutional Neural Network Architectures for Matching Natural Language Sentences[J]. International Conference on Neural Information Processing Systems, 2014,3:2042-2050
[19]	Bouvrie J. Notes on Convolutional Neural Networks[J]. Neural Nets, 2006: 38-44
[20]	Herreragomez A, Porter RM. Mixed linear-nonlinear least squares regression[J]. Mathematics, 2017: 49-65
[21]	Barham RH, Wanzer D. An Algorithm for Least Squares Estimation of Nonlinear Parameters When Some of the Parameters are linear[J]. Technometrics, 1972,14(3):757-766 doi: 10.1080/00401706.1972.10488964 URL

结合改进卷积神经网络和最小二乘法的颅骨性别鉴定

Skull sex identification using improved convolution neural network and least squares method

在线阅读

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 21

相关文章 2

编辑推荐

Metrics

本文评价

[1]	何欣龙, 梅宏成, 王继芬, 朱军, 郭鹏然, 胡灿, 何亚, 龚晓晓, 胡展纶. 新疆和田、哈密和阿勒泰市居民指甲中氢、氧稳定同位素的比较[J]. 人类学学报, 2021, 40(05): 888-894.
[2]	何欣龙, 梅宏成, 王继芬, 朱军, 刘昌景, 胡灿, 何亚, 张秀梅, 何玉录. 人指甲中氢、氧稳定同位素在法医学领域的应用[J]. 人类学学报, 2020, 39(03): 483-494.