基于线条形状和主方向墨色分布特征的中国画画家识别算法

发布时间：2018-01-23 04:33

本文关键词： 中国画分类方向梯度直方图主方向墨色线条形状特征提取　出处：《模式识别与人工智能》2017年10期 　论文类型：期刊论文

【摘要】：目前针对中国画的研究主要集中在基于内容的图像分析上,但是对于中国画的分类识别,更重要的是艺术风格而非内容.中国画的本质是笔墨运用,笔道的线条形状和墨色构成是鉴别艺术风格的重要因素.因此,文中提出基于笔墨特征的中国画画家识别算法.首先提取墨线的形状特征和墨色的布局特征,然后综合上述2种特征,作为支持向量机的输入训练得到画作分类器.实验表明,文中算法在平均查全率和查准率上较优,可以用于中国画的数字化分析、理解和识别,为中国画传承和鉴赏提供有效的数字工具.
[Abstract]:At present, the research on Chinese painting is mainly focused on content-based image analysis, but for the classification and recognition of Chinese painting, it is more important to the artistic style than the content. The essence of Chinese painting is the use of ink and brush. Line shape and ink color are important factors to distinguish the artistic style. Therefore, this paper proposes a recognition algorithm for Chinese painting painters based on brush and ink features. Firstly, the shape features of ink lines and the layout features of ink colors are extracted. Then synthesizing the above two features, as input training of support vector machine to get the painting classifier. Experiments show that the algorithm in the average recall and precision is better, can be used in the digital analysis of Chinese painting. Understanding and recognition, for the inheritance and appreciation of Chinese painting to provide an effective digital tools.
【作者单位】：天津财经大学理工学院;
【基金】：国家自然科学基金项目(No.61502331) 天津市自然科学基金项目(No.15JCQNJC00800,16JCYBJC42000) 中国民航信息技术科研基地开放课题(No.CAAC-ITRB-201504)资助~~
【分类号】：J212;TP391.41
【正文快照】： Supported by National Natural Science Foundation of China(No.61502331),Natural Science Foundation of Tianjin(No.15JCQNJC00800,16JCYBJC42000),Open Project of Information Technology Research Base of Civil Aviation Administration of China(No.CAAC-ITRB-20150

【相似文献】