Chinese painting and calligraphy image recognition technology based on pseudo linear directional diffusion equation

Abstract

Abstract In order to study the image recognition technology of traditional Chinese painting and calligraphy based on pseudo linear directional diffusion equation. Firstly, in the pixel domain, an overall style feature extraction algorithm of ink painting based on histogram is proposed, and then the local area with the most representative stroke style is located based on Sobel edge detection method to obtain the local detail style information describing the stroke. Then, using image processing, computer vision and other technologies, combined with some quantifiable authenticity identification indexes of Chinese painting, this paper designs a authenticity identification method of Chinese painting based on image recognition to assist the authenticity identification of Chinese painting and improve the reliability of authenticity identification. The experimental results show that the fusion algorithms based on information entropy and distance are proposed in this paper. On the premise of maximizing the artistic style, the fusion algorithm combines the advantages of the overall and local class results, so as to realize the automatic classification of ink painting based on artistic style and output the classification results.