Stroke Extraction Algorithm of Clerical Script in Han Dynasty Based on Contour: Take “Stele of Cao Quan” as an Example

• Published in: Mobile information systems Vol. 2022; pp. 1 - 10
• Main Authors: Guoqing, Liu, Changning, Hao, Jingbo, Yan, Jing, Dong, Zuolong, Zhao, Lujia, Hao
• Format: Journal Article
• Language: English
• Published: Amsterdam Hindawi 05.09.2022; John Wiley & Sons, Inc
• Subjects: Algorithms; Calligraphy; Contour matching; Cultural resources; Efficiency; Information systems; Inheritances; Inscriptions; Methods; Oxidation; Pattern recognition; Template matching; Thinning; Tracking; China
• ISSN: 1574-017X; 1875-905X; 1875-905X
• DOI: 10.1155/2022/5066994

Under natural conditions, the inscription will be damaged by corrosion and oxidation, which is very unfavorable for the inheritance of this cultural heritage. Therefore, how to protect and study ancient calligraphy inscriptions by modern scientific and technological methods is a scientific and technological problem before us. In this paper, taking “Stele of Cao Quan” as an example, a stroke extraction algorithm of clerical script in Han Dynasty based on contour is proposed. Starting from the Chinese character skeleton, the basic skeleton is extracted by parallel thinning algorithm, and the redundant points in the skeleton are eliminated by matching with the construction template. In addition, the contour tracking thinning algorithm is analyzed. Aiming at the problems that the starting point is difficult to determine and the tracking process is complicated, the method of determining the tracking starting point is given, and some problems existing in the tracking process are solved. The experimental results show that the accuracy of regular script, running script, and clerical script is 88.35%, 86.71%, and 84.09%, respectively. Experiments show that the contour-based clean script stroke extraction algorithm can correctly extract clean script strokes.