首页 > 论文 > 光谱学与光谱分析 > 39卷 > 3期(pp:756-764)

可见-近红外光谱的古陶瓷断代分类识别

Visible-Near Infrared Spectroscopy Based Chronological Classification and Identification of Ancient Ceramic

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

为客观、 有效对古陶瓷进行无损断代, 提出了一种基于可见-近红外光谱古陶瓷断代分类识别方法。 耀州窑古陶瓷跨代较多, 且不同朝代之间具有物理相似性, 因此耀州窑的断代具有一定的挑战性。 以耀州窑为研究对象, 在采用紫外-可见-近红外光谱分析仪获取古陶瓷不同朝代的多光谱数据的基础上, 提出基于分数阶微分对光谱数据进行预处理, 避免微分预处理常用的一阶微分和二阶微分遗漏中间过渡信息, 同时压制并消除光谱数据中的背景信息和噪声干扰。 实验结果表明, 未进行微分处理(0阶)时, 耀州窑不同年代古陶瓷的分类准确率仅为84.8%, 而基于不同分数阶微分的分类准确率均较0阶明显高, 分数阶微分的最优阶数为0.7阶。 另外, 提出基于深度信念网络对不同朝代古陶瓷进行断代分类, 首先采用层叠的受限玻尔兹曼机(RBM)对深度网络进行预训练, 提取光谱数据高层特征以消除光谱数据中的冗余特征。 实验结果表明, 光谱数据经RBM降维之前特征间的相关系数为0.885 7, 经第一层和第二层RBM降维后的相关系数分别为0.544 6和0.391 5, 特征间的相关性明显下降, 冗余度明显减少。 然后将RBM预训练得到的权值参数对BP神经网络进行初始化, 并对深度信念网络进行微调, 在克服BP神经网络因随机初始化权值参数而陷入局部最优局限性的同时, 提升网络训练主动性。 实验可得, 深度信念网络的最优RBM数量为2, RBM隐藏层最优节点数为100。 最后, 为避免小样本数据基于深度信念网络进行训练易出现过拟合, 提出了一种Dropout随机丢弃策略, 在深度信念网络训练阶段以一定概率随机让网络某些隐含层节点的权重不工作, 以减少网络训练过程特征之间的相互依赖性, 实验可得当Dropout丢弃比例为0.45时, 分类性能最高。 采用所提方法, 耀州窑不同朝代古陶瓷分类的平均准确率为93.5%, 而耀州窑五代时期的分类识别率最高为96.3%。 通过与同类古陶瓷断代分析方法的客观定量对比, 表明所提方法有效、 可行, 为古陶瓷的断代提供了新方法。

Abstract

Visible-near infrared spectroscopy based chronology classification of ancient ceramic method has been proposed to make the identification more objective and accurate. Yaozhou kiln exists in many dynasties and it has great similarity between different dynasties. Therefore, age identification of Yaozhou kiln faces great challenges. Taking Yaozhou kiln as the research object, some multi-spectral data of ancient ceramic from different dynasties are gotten from ultraviolet-visible near infrared spectroscopy analyzer. To avoid the first-order and second-order differential missing intermediate transition information, a fractional-order differential preprocessing method is proposed to suppress and eliminate the background information and noise from spectral data. The experimental results show that the classification accuracy of Yaozhou kiln in different dynasties is only 84.8% when the differential processing is not performed (0th order), while the classification accuracy based on different fractional differentials is obviously higher than that of 0th order. And the optimal order is 0.7. Then, a deep belief network based ancient ceramic classification method is proposed. First, stacked restricted Boltzmann machine (RBM) is employed to extract some high-level features during pre-training stage. The results show that the correlation coefficient between the features before RBM dimension reduction is 0.885 7, while the correlation coefficients after dimension reduction by the first and second RBM are 0.544 6 and 0.391 5 respectively, which means the redundancy is obviously cut back. Then some weight and bias values trained by RBM are used to initialize BP neural network. The whole deep belief network is fine-tuned by BP neural network to promote the initiative performance of network training and overcome local optimal limitation of the neural network due to the random initializing weight parameter. Experimentally, the optimal number of RBMs in depth belief network is 2, and the optimal number of RBM hidden layer units is 100. Meanwhile a dropout strategy is put forward to randomly ignore neurons of some hidden layers to reduce interdependence between features in the network training process and prevent over-fitting from some small data. When the ratio of Dropout is 0.45, the classification accuracy is highest. According to the method mentioned in this paper, the chronology classification accuracy in Yaozhou kiln is 93.5%, and accuracy of Yaozhou kiln in the Five Dynasties is highest, reaching 96.3%. Comparisons with some chronology classification methods highlight the superior performance of the developed method.

中国激光微信矩阵
补充资料

中图分类号:TP391.7

DOI:10.3964/j.issn.1000-0593(2019)03-0756-09

基金项目:国家自然科学基金项目(51232008, 51672302), 上海大学文化遗产保护基础科学研究院创新团队资助

收稿日期:2017-11-20

修改稿日期:2018-04-25

网络出版日期:--

作者单位    点击查看

吴晓萍:上海大学通信与信息工程学院, 上海 200444
管业鹏:上海大学通信与信息工程学院, 上海 200444新型显示技术及应用集成教育部重点实验室, 上海 200072
李伟东:中国科学院上海硅酸盐研究所, 上海 201899
罗宏杰:上海大学文化遗产保护基础科学研究院, 上海 200444

联系人作者:吴晓萍(wonderjx@t.shu.edu.cn)

备注:吴晓萍, 女, 1994年生, 上海大学通信与信息工程学院硕士研究生

【1】Legnaioli S, Garcia F A, Andreotti A, et al. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2013, 100(12): 144.

【2】LI Ding-ding, SHI Meng(李丁丁, 史 萌). Journal of Qufu Normal University·Natural Science(曲阜师范大学学报·自然科学版), 2015, 41(2): 50.

【3】Maritan L, Holakooei P, Mazzoli C. Applied Clay Science, 2015, 114(1): 540.

【4】WENG Zheng-kui, GUAN Ye-peng, LUO Hong-jie(翁政魁, 管业鹏, 罗宏杰). Journal of the Chinese Ceramic Society(硅酸盐学报), 2017, 45(12): 1833.

【5】WU Jun, YIN Li, ZHANG Mao-lin, et al(吴 隽, 尹 丽, 张茂林, 等). Journal of Ceramics(陶瓷学报), 2014, 35(4): 429.

【6】LIU Wei, LIU Chang-hong, ZHENG Lei(刘 伟, 刘长虹, 郑 磊). Transactions of the Chinese Society of Agricultural Engineering(农业工程学报), 2014, 30(10): 145.

【7】Liu N, Yang Y, Wang Y, et al. Ceramics International, 2017, 43(13): 10460.

【8】Zhu G F, Zhu H Q, Yang C H, et al. Journal of Optical Technology, 2017, 84(5): 355.

【9】ZHANG Shao-yang, XIE Yuan-yuan, ZHANG Xin(张绍阳, 解源源, 张 鑫). Optics and Precision Engineering(光学精密工程), 2014, 22(3): 779.

【10】MENG Xi, WANG Ke-jian, HAN Xian-zhon(孟 惜, 王克俭, 韩宪忠). Guizhou Agricultural Sciences(贵州农业科学), 2017, 45(10): 156.

【11】Xie C, Lv J, Li X. Soft Computing, 2016, 21(21): 6471.

【12】Fischer A, Igel C. Pattern Recognition, 2014, 47(1): 14.

【13】Zhang H, Gong M, Zhang P, et al. IEEE Geoscience and Remote Sensing Letters, 2016, 13(11): 1666.

引用该论文

WU Xiao-ping,GUAN Ye-peng,LI Wei-dong,LUO Hong-jie. Visible-Near Infrared Spectroscopy Based Chronological Classification and Identification of Ancient Ceramic[J]. Spectroscopy and Spectral Analysis, 2019, 39(3): 756-764

吴晓萍,管业鹏,李伟东,罗宏杰. 可见-近红外光谱的古陶瓷断代分类识别[J]. 光谱学与光谱分析, 2019, 39(3): 756-764

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF