Welcome to Francis Academic Press

Academic Journal of Materials & Chemistry, 2023, 4(2); doi: 10.25236/AJMC.2023.040201.

A study of the composition of ancient glass based on a principal component analysis model


Tianyu Zhu

Corresponding Author:
Tianyu Zhu

College of Physics and Electromechanical Engineering, Hubei University of Education, Wuhan, China


In the early years of the country, China used the Silk Road to trade glass, learn how to make it and make it. As a result, the glass made in China was similar in appearance to exotic glass, but differed in chemical composition. The main objective of this paper is to develop a compositional analysis and identification model for ancient glass objects, which can be used to study the degree of surface weathering of glass objects and its relationship to glass type, decoration and color, as well as the types of glass objects that belong to unknown categories. The chemical composition content of the pre-weathering artifacts was further predicted and the correlation between the chemical components was analyzed. This paper introduces a glass classification method based on principal component analysis and a Gaussian mixture model with the expectation maximization principle for the identification of glass artifacts. Then a confusion matrix was introduced to perform sensitivity analysis on the model, and the correct and recall rates were 0.911 and 0.844, respectively, with good scores.


principal component analysis, gaussian mixture model, glass classification

Cite This Paper

Tianyu Zhu. A study of the composition of ancient glass based on a principal component analysis model. Academic Journal of Materials & Chemistry (2023) Vol. 4, Issue 2: 1-5. https://doi.org/10.25236/AJMC.2023.040201.


[1] Wu B., Yu D. Deep convolutional neural network-based classification and detection of cell phone glass cover surface defects [J]. Software Engineering, 2021, 24(12):6-10.

[2] Zhang Jian, Hao Nannan. Gemological classification and identification of meteorite glass[J]. Shandong Chemical Industry, 2021, 50(17):160-161.

[3] Liu C. X., Deng P. F., Pan S. H., Luo W. D., Zhang S. L., Li T. M.. Classification and research status of fireproof glass [J]. Guangzhou Chemical Industry, 2021, 49(15):16-18.

[4] Yang X. Detection and classification of glass defects based on machine vision[D]. Fujian College of Engineering, 2021.

[5] Li Yingli. Research on glassware excavated in China from the 10th to the 12th century[D]. Liaoning Normal University, 2021.

[6] Ren Zhanzhan. Research on glassware of Wei, Jin, North and South Dynasties, Sui and Tang [D]. Suzhou University, 2020.

[7] Huang Jianyi. Research on the detection and classification method of cell phone panel glass defects based on digital image processing [D]. Xi'an University of Electronic Science and Technology, 2020.

[8] Li Xiangjing. Study on the correlation between the proportion of solid component volume in pulmonary ground glass shadow and the pathological classification of lung adenocarcinoma[D]. Henan University, 2020.

[9] Kong Yang. Study of glass precrystallization behavior of fluoroaluminosilicate system[D]. Qilu University of Technology, 2019.

[10] Du Wenxiu, Yang Juan, Gao Shuai, Wu Benhua, Wang Yinmao, Gao Guozhong. Introduction to the characteristics and classification of fireproof glass [J]. Anhui Architecture,2015,22(04):179-181.

[11] He Yuanyuan. Research on contemporary decorative art glass design in China [D]. Wuhan University of Technology,2015.

[12] Huang Liguang, Chen Jinfang. Classification and status of electric glass melting kilns [J]. Glass, 2012, 39(12):37-39.

[13] Sun H, Liu M, Li L, et al. A new classification method of ancient Chinese ceramics based on machine learning and component analysis [J]. Ceramics International, 2019, 46(6).

[14] Daniel Ramos, Grzegorz Zadora. Information-theoretic feature selection using data obtained by scanning electron microscopy with energy dispersive X-ray spectroscopy for the classification of glass traces [J]. Analytica Chimica Acta, 2011, 705(1).

[15] Zetlein Michael B. Glass classification by elemental composition. A numerical assessment [J]. Forensic Science International, 1979, 13.