透過您的圖書館登入
IP:18.191.84.32
  • 學位論文

資訊科學在安陽出土甲骨拼合上的應用

Exploiting Information Science Technology to Rejoin An-yang Oracle Bones/Shells

指導教授 : 陳光祖 顧坤惠
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


利用科技方法還原或重建考古遺物是考古學上一個近來備受注目的課題,而近代利用科學方法發掘所獲得的甲骨不但是重要的考古遺物,將碎裂分散的甲骨綴合復原也是解讀甲骨文必要的工作,具有文史等方面的多重意義。鑑於甲骨綴合一直被認為是耗時甚大的工程,本研究提出以資訊科學中影像處理及拼合演算法等技術,將甲骨轉換為多邊形向量以模擬甲骨邊緣的拼合,期能對甲骨拼合工作做出實用的建議,克服材料分散各地不易實地綴合的困難,也減少人為拼合過程中對實物可能造成的損害。本文提出合適的影像處理技巧以及數種拼合演算法,並實際撰寫程式,根據中央研究院歷史語言研究所典藏甲骨之數位圖檔,對中國河南安陽殷墟出土之商代甲骨碎片進行拼合測試。於得出可能拼合的候選組合後,除標示拼合面之外並對每一組合加以評分,以排列檢驗時的優先順序。實驗結果顯示本系統可以達成80%∼90%的準確率,使用者也可依需要選擇不同的組合進行檢驗。此結果顯示系統對於未來甲骨拼合工作可提出有益的建議,可望提供日後學者綴合時的參考。最終則希望能擴充至國內外其它的甲骨資料庫,與其它種類之有字遺物如簡牘、帛書,以及銅器、陶瓷器等遺物的拼合,達到更廣泛性的實用目的。同時,藉由本文也嘗試考古與資訊科學間跨學科整合的方法與可行性。

並列摘要


Applying scientific and technological methods to reconstructing archaeological remains has become a very important issue in the modern archaeology research. At the same time, though being a must-done work before deciphering the inscriptions, rejoining oracle bone fragments manually has been regarded as a time consuming and tedious job. This thesis proposes an inter-disciplinary approach to exploit information science technologies to automatically piece broken oracle bones. The proposed approach integrates the image processing tools and several different assembling algorithms to recommend possible matches. The goal is to overcome the difficulty of access of the artifacts from different collections and to minimize the possible damage caused to the artifacts through manual assembling. The algorithm is implemented and tested on pictures of An-yang oracle bones from Institute of History and Philology, Academia Sinica. It is capable of ranking the matched bones based on some designed scores. The promising experiment results show that the system can reach 80~90% accuracy. The ultimate goal is to provide practical recommendations for not only oracle bone but also other archaeological materials. This thesis also demonstrates how an interdisciplinary approach can be designed and applied with the hope to advance the collaboration in between computer science and archaeology researches.

參考文獻


國立成功大學圖書館
1998 Solving Jigsaw Puzzles by a Robot. IEEE Trans. Robotics and Automation 5(6):752-764.
Canny, John F.
1986 A Computational Approach to Edge Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence vol. 8:679-698.
Chou, Hung-hsiang

延伸閱讀