Recognition of Oracle Bone Inscriptions by Clustering and Matching on the Hough Space

DOI Web Site
  • Meng Lin
    College of Science and Engineering, Ritsumeikan University
  • Izumi Tomonori
    College of Science and Engineering, Ritsumeikan University
  • Oyanagi Shigeru
    College of Information Science and Engineering, Ritsumeikan University

Bibliographic Information

Other Title
  • ハフ空間上でのクラスタリングとマッチングによる甲骨文字の認識
  • ハフ クウカン ジョウ デ ノ クラスタリング ト マッチング ニ ヨル コウコツブンジ ノ ニンシキ

Search this article

Description

Oracle bone inscriptions (OBIs) are one of the oldest characters which were inscribed on the bone of cattle or turtle shells in the period of Shang. Recognizing the OBI is important for understanding the origin of characters, history research, etc. However, aged caused OBI to be less legible. The OBI was inscribed by the sharp objects. Hence, the structure of OBI is simple and constructed mainly by lines. Our approach extracts the line features by Hough transform, and recognizes the OBI from templates. First, line candidates are estimated using voting scheme of the Hough transform. Second, the lines and the number of the lines are estimated by clustering neighbor of line candidates. The recognition of OBI is performed by minimizing the distance between the corresponding points on the Hough space of the target and template images. The experimental results show that about 87% OBIs are recognized in 7 kinds, 52 pieces of OBIs with 11 kinds of templates.

Journal

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top