Research on Dynamic Ink and Character Skeleton Extraction of Calligraphic Style in Calligraphy Creation

With the development of computer technology, the study of dynamic inking and glyph skeleton extraction of font style in calligraphy creation is changing day by day. In this paper, on the basis of image feature extraction by convolutional neural network, particle system is utilized to capture the dynamic ink mark of fonts in calligraphy creation, and the character skeleton of Chinese characters is extracted based on adversarial neural network, and finally, twin neural network is utilized to classify the style of the acquired dynamic ink mark of fonts and character skeleton. Through the research and analysis, it is found that the algorithm recognition accuracy can be increased to 95% by using ReLU function as the activation function of the convolutional layer, and the generative adversarial network with the introduction of contrast loss and Haar wavelet analysis can alsdo reach more than 90% accuracy in glyph skeleton extraction. In addition, the accuracy rate of five style classifications for all fonts also remains above 80%. The study of dynamic ink and glyph skeleton extraction of font styles in calligraphy creation using different techniques will greatly enrich the research and analysis of Chinese calligraphy.