首页> 外文期刊>Optik: Zeitschrift fur Licht- und Elektronenoptik: = Journal for Light-and Electronoptic >Correlation and image moment approaches to analyze the Glagolitic script carved in stone tablets
【24h】

Correlation and image moment approaches to analyze the Glagolitic script carved in stone tablets

机译:相关性和图像矩分析法分析石碑刻刻的早古文字

获取原文
获取原文并翻译 | 示例
       

摘要

In relatively recent paleographic studies, a new type of the Glagolitic script, namely triangular type, was introduced as being older than the rounded type (commonly considered as the oldest known type) with possible mixing of letter types in the early period of Glagolitic writing. That hypothesis, drawn from the so-called generative model, has been lacking of a quantitative analysis. Here, we report on such an analysis in which the forms of the original letters carved in two among oldest stone tablets are compared to the models of the triangular and rounded letters. To calculate similarity measures, two different approaches were used. In the first approach, we use the standard correlator systems that are fast but sensitive to various letter deformations. The second approach, via calculating Fourier-Jacobi image moments, is rather slow but possesses invariance to input image deformations. To accelerate the entire recognition and classification process of the later approach, we introduced so-called sensitivity vector consisting of those moments that are robust to rotational and scale change deformations of input image. Although different in details, both approaches indicate presence of a mixture of both letter types supporting thus one of the hypothesis claims.
机译:在相对较近的古地理研究中,引入了一种新型的古拉格利特文字,即三角形文字,它比圆角文字(通常被认为是最古老的已知文字)更旧,并且在古拉格利特文字的早期可能混合了字母类型。从所谓的生成模型得出的假设一直缺乏定量分析。在这里,我们报告了这样一种分析,其中将刻在两个最古老的石碑中的原始字母的形式与三角形和圆形字母的模型进行比较。为了计算相似性度量,使用了两种不同的方法。在第一种方法中,我们使用快速但对各种字母变形敏感的标准相关器系统。第二种方法是通过计算傅里叶-雅各比影像矩,虽然速度较慢,但​​具有输入影像变形的不变性。为了加快后一种方法的整个识别和分类过程,我们引入了所谓的敏感度矢量,其中包括对输入图像的旋转和比例变化变形具有鲁棒性的那些力矩。尽管在细节上有所不同,但这两种方法都表明存在两种字母类型的混合物,从而支持了一项假设要求。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号