Research And Application Of Diversity Expression Methods For Tibetan Classic Literature Fonts

Since the creation of Tibetan characters in the 7th century,Tibetan compatriots have recorded their own culture in their own words and written a large number of classic documents,providing valuable resources for future generations to study Tibetan history,religion,culture,etc.Among them,Tibetan Scripture is the most complete document to preserve the original appearance of Buddhism in the world,so the study of Tibetan Scripture is of great significanceAt present,the research on the Tibetan classic literature mainly focuses on the content recognition and image processing,and does not pay attention to the diversity and beauty of the font in the original engraving of the Sutra The realization of font diversity expression can not only help to fit the original appearance of Tibetan classic literature,but also provide ideas and help for the preservation and inheritance of our excellent culture,which has far-reaching practical significance.In this paper,the diversity of font expression as the goal,focusing on the unique font in the Tibetan Scripture,the main work is as follows:(1)This paper expounds the subject of font diversity expression.Based on the analysis of the current research status of Tibetan Classical Literature and various and unique carved fonts,this paper describes the topic of font diversity expression,and gives the basic composition modules and basic ideas to achieve font diversity expression(2)Image processing and feature extraction of the original Tibetan Scripture image.The component library of 170 components which can form 7240 characters is determined.Starting from the actual situation of the original image of Tibetan Scripture,the steps of image processing are proposed,and the first to the tenth segments of the original image of Scripture are used as image processing materials to get a single component image of each component.The feature matrix of each single component image is obtained by multi projection(3)The feature matrix of components is processed by clustering.The improved k-center algorithm is used to select the feature matrix with the largest distance as the clustering center,and then the number of representative Fonts is determined.Using the feature matrix as the material for clustering,K representative fonts of each component are obtained,and then K representative fonts are obtained(4)Through the replacement algorithm,we get the effect of font diversity.After clustering,the number of fonts in each cluster is counted,the font model of basic font+discrete font is established,and the ratio of basic font to discrete font is calculated to select from basic font and discrete font.When selecting the discrete fonts,the random number generated by Roulette is used to get the font number.According to the replacement algorithm,the problem of selecting the representative Fonts is solved,and the effect of font diversity is achieved by using the representative fonts.