Historical stylistic evolution of traditional temple murals based on time series analysis

Abstract

Abstract In this paper, firstly, on the basis of time series analysis, considering the validity of data research data, it is necessary to carry out the pre-processing process of independent variables and dependent variables of temple frescoes and then select the characteristic variables of temple frescoes, and get the multivariate time series of temple frescoes through wavelet correlation analysis of the characteristic variables of temple frescoes. Then, construct the classification model of temple frescoes based on the multivariate time series through the image processing of temple frescoes, get the feature values applicable to the model classification, considering the accuracy and convergence of the model classification, the temple frescoes features are needed to be extracted and trained, and at the same time, the evolution of the historical style of the traditional temple frescoes is analyzed by examples. The results show that in terms of performance, the method of this paper achieves 90.34% accuracy, and the comparison results of two indicators, recall and F1 value, also show that this paper shows good performance compared with the other four models. In the correlation analysis, it is found that there is a significant correlation between the rich artistic heritage of dynasties (P=0.66>0.05), the economic prosperity of dynasties (P=0.75>0.05) and the evolution of the historical style of temple frescoes, and it is proposed that traditional temple frescoes are protected and developed paths.