基于迭代加权图像的藏文古籍逐级分类仿真  

Hierarchical classification Simulation of Tibetan Ancient Books Based on Iterative Weighted Image

作　　者：沈淑涛 高飞 梁家瑞 SHEN Shu-tao;GAO Fei;LIANG Jia-rui(Tibet University School of Information Science and Technology,Xizang Lasa 850000,China;Taiyuan University of Technology,School of Software,shanxi yuci 030600,China)

机构地区：[1]西藏大学,信息科学技术学院,西藏拉萨850000 [2]太原理工大学,软件学院,山西榆次030600

出　　处：《计算机仿真》2021年第7期410-414,共5页Computer Simulation

基　　金：2018年度西藏大学科研培育基金项目(成长计划)(ZDCZJH18-15);西藏自治区教育厅‘计算机及藏文信息技术国家级团队和重点实验室建设’(藏教财指[2018]81号);。

摘　　要：为解决传统藏文古籍分类不准确和分类效率低的问题,提出基于迭代加权图像的藏文古籍逐级分类方法。首先通过迭代加权算法对藏文古籍进行样本建造,对样本进行反复协方差矩阵训练,使后续藏文古籍逐级分类更为精准。然后使用图像信息特征提取方法,对藏文古籍迭代加权图像进行特征信息提取,挑选出能为逐级分类提供较大贡献的图像信息。最后通过迭代加权图像文本表示对藏文古籍进行逐级分类。仿真结果证明,所提方法在藏文古籍分类中,有着分类准确和分类效率高等优势。In order to solve the problems of inaccurate and inefficient classification of traditional Tibetan ancient books,a hierarchical classification method of Tibetan ancient books based on iterative weighted images was proposed.Firstly,the iterative weighting algorithm was used to construct the samples of Tibetan ancient books,and then the samples were trained repeatedly by covariance matrix,so that the subsequent classification of Tibetan ancient books was more accurate.After that,the image information feature extraction method was used to extract the feature information from iterative weighted images of ancient Tibetan books,and the image information that can provide greater contribution to the hierarchical classification was selected.Finally,the Tibetan ancient books were classified by text representation of iteration weighted image.Simulation results show that the proposed method has high accuracy and high efficiency in the classification of Tibetan ancient books.

关 键 词：藏文古籍 迭代加权 逐级分类 协方差矩阵 

分 类 号：TU92[建筑科学—建筑设计及理论]