Digital Inheritance of POPF Based on Image Database and Identification Model

Pecking Opera Painted Faces (POPF) is a Chinese intangible cultural heritage full of aesthetic value. However, their cultural connotations become less known in modern society. It is therefor, necessary to make public easily understand unique cultural connotation and aesthetic value of POPF. In this project, we built a POPF image database based on the image classification model, with the aim to better showcase POPF. The images are classified into multiple categories according to the traditional connotation patterns. By using the MLP (Multilayer Perceptrons) algorithm classifier, the classification accuracy of model has approached 70\%. Using those tools, the uncertainty of information about images of POPF can be reduced appreciably, and would benefit the innovation of derivatives around POPF.