计算机工程与应用 ›› 2022, Vol. 58 ›› Issue (14): 194-199.DOI: 10.3778/j.issn.1002-8331.2011-0118

• 图形图像处理 • 上一篇    下一篇

基于知识推理的纹饰演化关系发现

卢明,侯小刚,韩晓彤,赵海英   

  1. 1.甘肃省科学院 自动化研究所,兰州 730030
    2.北京邮电大学 人工智能学院,北京 100876
    3.北京邮电大学 数字媒体与设计艺术学院,北京 100876
    4.北京邮电大学 世纪学院 移动媒体与文化计算北京市重点实验室,北京 102101
  • 出版日期:2022-07-15 发布日期:2022-07-15

Discovery of Evolutionary Relationship of Ornaments Based on Knowledge Reasoning

LU Ming, HOU Xiaogang, HAN Xiaotong, ZHAO Haiying   

  1. 1.Institute of Automation of Gansu Academy of Sciences, Lanzhou 730000, China
    2.School of Artificial Intelligence, Beijing University of Posts and Telecommunications, Beijing 100876, China
    3.School of Digital Media and Design Arts, Beijing University of Posts and Telecommunications, Beijing 100876, China
    4.Beijing Key Laboratory of Mobile Media and Cultural Computing, Century College, Beijing University of Posts and Telecommunications, Beijing 102101, China
  • Online:2022-07-15 Published:2022-07-15

摘要: 通过对传统纹饰的时间、空间、历史等诸多要素的梳理和分析,挖掘其时空演化规律,并探寻一套传承纹饰文化内涵的脉络。提取传统纹饰的视觉特征,以视觉相似度作为纹饰相关度的度量值,并依据纹饰相关属性构建传统纹饰语义网。利用所提的基于知识推理的演化关系发现算法预测传统纹饰的演化过程。以《中国纹样全集》为实验数据,抽取纹饰属性及其关系,整理出不同时期不同地域的纹饰共有1?500个,纹饰之间形成的关系有13?600个,通过实验验证了基于知识推理的传统纹样演化关系发现算法的可行性,并通过可视化呈现更直观地展示了纹饰的演化过程。

关键词: 传统纹饰, 演化关系发现, 纹饰语义网, 纹饰相关度

Abstract: Through sorting out and analyzing the element in time, space, history and other elements of traditional ornamentation, this paper exploits the rule of its spatio-temporal evolution, and explores a set of veins for inheriting the cultural connotation of ornamentation. In this paper, visual features of traditional ornamentation are firstly extracted, and the visual similarity is taken as the measurement value of the relevancy of ornamentation, and the semantic network of traditional ornamentation is constructed according to the related attributes of ornamentation. Then the evolutionary process of traditional ornamentation is predicted by the evolutionary relationship discovery algorithm based on knowledge reasoning. Finally, taking the Complete Set of Chinese Patterns as experimental data, the attributes and relationships of patterns are extracted, and a total of 1 500 patterns in different periods and different regions are sorted out, with 13 600 relationships formed between patterns. Through experiments, the feasibility of the traditional algorithm based on knowledge reasoning to discover the evolution relationship of patterns is verified, and the evolution process of patterns is shown more intuitively through visual presentation.

Key words: traditional ornamentation, discovery of evolutionary relations, decorative semantic network, relevance of decorative