甲骨文识别技术研究现状与展望
|
刘洋,助理教授,博士,E-mail:yang.liu27@whu.edu.cn |
|
陆逸,本科生 |
|
魏钰驰,本科生 |
|
孙智莹,本科生 |
|
朱立芳,讲师,博士 |
Copy editor: 刘远颖
收稿日期: 2023-01-28
网络出版日期: 2025-04-28
基金资助
国家自然科学基金青年项目“突发公共卫生事件公众心理应激信息表征及干预机制研究”(72204190)
教育部人文社科项目青年项目“基于社交机器人的突发公共卫生事件公众心理应激干预研究”(22YJCZH114)
中国博士后面上基金“突发公共卫生事件公众心理应激信息表征及干预机制研究”(2022M722476)
Research Status and Prospect of Oracle Bone Inscription Recognition Technology
Received date: 2023-01-28
Online published: 2025-04-28
[目的/意义]对数字人文视域下甲骨文识别研究进行系统性综述,为后续研究提供参考和借鉴,推动数字人文研究有效发展与古籍文字识别利用。[方法/过程]采用文献计量分析的方法,在WOS、中国知网等多个学术平台检索文献,共筛选103篇英文文献和52篇中文文献进行综述。[结果/结论]从传统识别技术、机器学习和深度学习3个层面解读甲骨文识别研究现状,但并未深入阐述识别算法机制。甲骨文识别技术由传统的特征提取逐渐转为基于深度学习的识别技术,在识别精度等方面有很大提升,但仍存在一些不足,同时甲骨文知识库、知识图谱的构建与领域知识的建立在该领域有较好的发展潜力。
刘洋 , 陆逸 , 魏钰驰 , 孙智莹 , 朱立芳 . 甲骨文识别技术研究现状与展望[J]. 知识管理论坛, 2023 , 8(2) : 115 -125 . DOI: 10.13266/j.issn.2095-5472.2023.010
[Purpose/Significance] Digital humanities research is a prominent research hotspot in the current academic circle. This study systematically reviewed the frontier research on oracle bone inscription recognition from the perspective of digital humanities, which provided reference for follow-up research, promoting the effective development of digital humanities research and the recognition and utilization of characters in ancient books. [Method/Process] The literature was retrieved from multiple academic platforms such as WOS and CNKI using the method of bibliometric analysis, and a total of 103 English literature and 52 Chinese literature were screened for review. [Result/Conclusion] Interpreting the research status of oracle bone inscription recognition from three levels: traditional recognition technology, machine learning and deep learning, which analyzed the research development process, and discussed the future development trend. This paper mainly conducted a systematic review of oracle bone inscription recognition research from the perspective of digital humanities, which analyzed existing research technologies and research directions, but did not elaborate on the recognition algorithm mechanism in depth. Oracle recognition technology has gradually changed from traditional feature extraction to deep learning-based recognition technology. Although the recognition accuracy has been improved, there are still shortcomings such as serious overfitting and low recognition efficiency. Meanwhile, the construction of oracle knowledge base and knowledge graph, and the establishment of domain knowledge have good development potential in this field.
Key words: digital humanities; oracle bone recognition; research progress; review
表1 用于深度学习的拓片图像识别数据统计表 |
| 作者 | 数据集 | 技术方法 | 结果 |
| 陈婷珠、吴少腾等[15] | 《殷墟小屯村中村南甲骨》:515片甲骨,6 230张甲骨单字图像 | 将甲骨文图像转化为编码 | 训练集:100% |
| 刘芳、李华飙等[16] | 《甲骨文合集》:4 378张甲骨文单字图像 | Mask R-CNN | 检测和识别准确率均达到95% |
| 闫升、刘芳等[17] | 中国国家博物馆馆藏甲骨精品拓片图像以及《甲骨文合集》中部分甲骨拓片图像和入《甲骨文常用字字典》辅助数据集 | 改进Mask R-CNN,实现检测与识别一体化 | 训练集:99.5% 测试集:61.7% |
| 林小渝、陈善雄等[18] | HCL2000数据集 | 甲骨文偏旁:BN-LeNet网络; 甲骨文合体字:OraNet模型 | 甲骨文偏旁:96.24%;甲骨文合体字:98.58% |
| 张颐康、张恒等[19] | 安阳师范学院甲骨文信息处理实验室甲骨文数据集(共295 466个样本),选取241类拓片甲骨文字样本 | 跨模态深度 度量学习 | 已知:86.7%;新类:62.1% |
| L. Meng、N. Kamitoku等[20] | 由“上海博物馆甲骨文字”扫描而来 | 自上向下扩展聚类(Top-Down Extension Clustering, TDE-C) 和依赖矩阵(Dependency Matrix, DM) | 首次使用深度学习识别真实甲骨文字符,准确率达到92.3% |
| L. Meng、B. Lyu等[21] | 一个由真实摩擦图像组成的甲骨文数据集(同类中的第一个数据集) | 单侧多箱检测器(Single Shot MultiBox Detector, SSD) | 准确率达到95% |
| L. Meng、N. Kamitoku等[22] | 由“上海博物馆甲骨文字”扫描而来 | SSD | 准确率达到97% |
| N. Wang、Q. Sun等[23] | 公开网络数据集 “殷契文渊” | YOLOv4模型(You Only Look Once version4) | 识别准确率达到75%,召回率达到90% |
| B. Du、G. Liu、W. Ge[24] | 公开网络数据集 “殷契文渊” | 双分支自我监督深度学习 | |
| X. Yue、B. Lyu等[25] | 立命馆大学白川静香东亚文字文化研究所 “白川字体” | 动态K-means聚类、定向梯度直方图(Histogram of oriented gradient, HOG)特征、神经网络 | 区分噪声和字符的准确率达到96.5%,字符分类准确率达到74.91% |
| C. S. Zhang、R. X. Zong等[26] | 真实甲骨文数据集OB-Rejoin、甲骨文注释数据集OracleBone-8000 | 甲骨文重联算法、基于深度学习的场景文本检测算法、深度模型匹配算法 | 碎片匹配前10%准确率为98.39%;甲骨文定位F得分为89.7%;甲骨文识别总体准确率为80.9% |
刘洋:确定选题,提出研究思路,修改论文;
陆逸:分析和处理数据,撰写论文;
魏钰驰:分析和处理数据,撰写论文;
孙智莹:分析和处理数据,撰写论文;
朱立芳:修改论文。
| [1] |
沃尔什,科布,弗雷默里,等.iSchool中的数字人文[J].陈怡,译.数字人文研究,2021,1(3):93-112.
|
| [2] |
邓君,王阮.数字人文视域下口述历史档案资源知识发现模型构建[J].档案学研究,2022(1):110-116.
|
| [3] |
李巧明,王晓光.跨学科视角下数字人文研究中心的组织与运作[J].数字图书馆论坛,2013(3):26-31.
|
| [4] |
陈力.数字人文视域下的古籍数字化与古典知识库建设问题[J].中国图书馆学报,2022,48(2):36-46.
|
| [5] |
刘乾先,董莲池,张玉春,等.中华文明实录[M].哈尔滨:黑龙江人民出版社,2002.
|
| [6] |
卢芯怡.新时期甲骨文应用研究述评[J].汉字文化,2020(21):73-78.
|
| [7] |
刘国英.基于深度学习的甲骨文字检测与识别[J].殷都学刊,2020,41(3):54-59.
|
| [8] |
李锋,周新伦.甲骨文自动识别的图论方法[J].电子科学学刊,1996(S1):41-47.
|
| [9] |
周新伦,李锋,华星城,等.甲骨文计算机识别方法研究[J].复旦学报(自然科学版),1996(5):481-486.
|
| [10] |
吕肖庆,李沫楠,蔡凯伟, 等.一种基于图形识别的甲骨文分类方法[J].北京信息科技大学学报(自然科学版),2010,25(S2):92-96.
|
| [11] |
顾绍通.基于拓扑配准的甲骨文字形识别方法[J].计算机与数字工程,2016,44(10):2001-2006.
|
| [12] |
CRISTIANINI N,TAYLOR J S.支持向量机导论[M]. 李国正,王猛,曾华军,译.北京:电子工业出版社,2004.
|
| [13] |
SHI X. Research on oracle word structure analysis based on support vector machine[D]. Shanghai: East China Normal University, 2010.
|
| [14] |
LIU Y, LIU G. Oracle-bone inscription recognition based on svm[J]. Journal of Anyang Normal University, 2017,2:54-56.
|
| [15] |
陈婷珠,吴少腾,吴江,等.基于编码的甲骨文识别技术研究[J].中国文字研究,2019(1):1-12.
|
| [16] |
刘芳,李华飙,马晋,等.基于Mask R-CNN的甲骨文拓片的自动检测与识别研究[J].数据分析与知识发现,2021,5(12):88-97.
|
| [17] |
闫升,刘芳,孙岱萌,等.博物馆基于人工智能的甲骨文知识普及与活化传承[J].中国博物馆,2021(3):110-116,144.
|
| [18] |
林小渝,陈善雄,高未泽,等.基于深度学习的甲骨文偏旁与合体字的识别研究[J].南京师大学报(自然科学版),2021,44(2):104-116.
|
| [19] |
张颐康,张恒,刘永革,等.基于跨模态深度度量学习的甲骨文字识别[J].自动化学报,2021,47(4):791-800.
|
| [20] |
Meng L, Kamitoku N, Yamazaki K. Recognition of oracle bone inscriptions using deep learning based on data augmentation[C]//2018 metrology for archaeology and cultural heritage (MetroArchaeo). Piscataway: IEEE, 2018: 33-38.
|
| [21] |
Meng L, Lyu B, Zhang Z, et al. Oracle bone inscription detector based on SSD[C]//International conference on image analysis and processing. Berlin: Springer, 2019: 126-136.
|
| [22] |
Meng L, Kamitoku N, Kong X, et al. Deep learning based ancient literature recognition and preservation[C]//2019 58th annual conference of the Society of Instrument and Control Engineers of Japan (SICE). Piscataway: IEEE, 2019: 473-476.
|
| [23] |
Wang N, Sun Q, Jiao Q, et al. Oracle bone inscriptions detection in rubbings based on deep learning[C]//2020 IEEE 9th joint international information technology and artificial intelligence conference (ITAIC). Piscataway: IEEE, 2020: 1671-1674.
|
| [24] |
Du B, Liu G, Ge W. Deep self-supervised learning for Oracle bone inscriptions features representation[C]//2021 IEEE 4th international conference on information systems and computer aided education (ICISCAE). Piscataway: IEEE, 2021: 7-11.
|
| [25] |
Yue X, Lyu B, Li H, et al. Deep learning and image processing combined organization of Shirakawa’s hand-notated documents on OBI research[C]//2021 IEEE international conference on networking, sensing and control (ICNSC). Piscataway: IEEE, 2021: 1-6.
|
| [26] |
Zhang C, Zong R, Cao S, et al. AI-powered oracle bone inscriptions recognition and fragments rejoining[C]//Proceedings of the Twenty-Ninth International Conference on International Joint Conferences on Artificial Intelligence, Yokohama, 2021: 5309-5311.
|
| [27] |
Liu Z, Wang X, Yang C, et al. Oracle character detection based on improved faster R-CNN[C]//2021 international conference on intelligent transportation, big data & smart city (ICITBS). Piscataway: IEEE, 2021: 697-700.
|
| [28] |
Liu W, Anguelov D, Erhan D, et al. Ssd: Single shot multibox detector[C]//Computer Vision–ECCV 2016: 14th European Conference, Amsterdam, The Netherlands, October 11–14, 2016, Proceedings, Part I 14. Springer International Publishing, 2016: 21-37.
|
| [29] |
林小渝. 基于深度学习的甲骨文偏旁与合体字识别的研究与实现[D]. 重庆: 西南大学,2021.
|
| [30] |
Guo Z, Zhou Z, Liu B, et al. An improved neural network model based on inception-v3 for Oracle bone inscription character recognition[J/OL]. Scientific programming, 2022[2023-01-27]. https://doi.org/10.1155/2022/7490363.
|
| [31] |
Fujikawa Y, Li H, Yue X, et al. Recognition of oracle bone inscriptions by using two deep learning models[J/OL]. International journal of dental hygiene, 2022[2023-01-27]. https://doi.org/10.1007/s42803-022-00044-9.
|
| [32] |
Guo J, Wang C H, Roman-Rangel E, et al. Building hierarchical representations for oracle character and sketch recognition[J]. IEEE transactions on image processing, 2016, 25(1): 104−118.
|
| [33] |
Gao F, Zhang J, Liu Y, et al. Image translation for Oracle bone character interpretation[J]. Symmetry, 2022, 14(4): 743.
|
| [34] |
Han W, Ren X, Lin H, et al. Self-supervised learning of orc-bert augmentator for recognizing few-shot oracle characters[C]//Proceedings of the Asian conference on computer vision, Kyoto: Revised Selected Papers, 2020: 652-668.
|
| [35] |
Li J, Wang Q F, Zhang R, et al. Mix-up augmentation for oracle character recognition with imbalanced data distribution[C]//Document analysis and recognition–ICDAR 2021: 16th international conference. Berlin: Springer International Publishing, 2021: 237-251.
|
| [36] |
Dazheng L. Random polygon cover for Oracle bone character recognition[C]//2021 5th international conference on computer science and artificial intelligence. New York: Association for Computing Machinery, 2021: 138-142.
|
| [37] |
Gao J, Liang X. Distinguishing oracle variants based on the isomorphism and symmetry invariances of oracle-bone inscriptions[J]. IEEE access, 2020, 8: 152258-152275.
|
| [38] |
Liu G, Ge W, Du B. Recognition of OBIC's variants by using deep neural networks and spectral clustering[C]//2021 IEEE 4th international conference on information systems and computer aided education (ICISCAE). Piscataway: IEEE, 2021: 39-42.
|
| [39] |
杨琳.数字化古典文献综述[J].中国史研究动态,2004(4):20-27.
|
| [40] |
门艺.由甲骨学工具书的编纂到甲骨文数据库的建设[J].漯河职业技术学院学报,2019,18(5):1-7.
|
| [41] |
栗青生,吴琴霞,王蕾.基于甲骨文字形动态描述库的甲骨文输入方法[J].中文信息学报,2012,26(4):28-33.
|
| [42] |
栗青生,吴琴霞,杨玉星.甲骨文字形动态描述库及其字形生成技术研究[J].北京大学学报(自然科学版),2013,49(1):61-67.
|
| [43] |
门艺,张重生.基于人工智能的甲骨文识别技术与字形数据库构建[J].中国文字研究,2021(1):9-16.
|
| [44] |
Huang S, Wang H, Liu Y, et al. Obc306: a large-scale oracle bone character recognition dataset[C]//2019 international conference on document analysis and recognition (ICDAR). Piscataway: IEEE, 2019: 681-688.
|
| [45] |
Xian-jin S H I, Shuang C A O, Chong-sheng Z, et al. Research on automatic annotation algorithm for character-level Oracle-bone images based on anchor points[J]. Acta electonica SINICA, 2021, 49(10): 2020-2031.
|
| [46] |
Xiong J, Liu G, Liu Y, et al. Oracle bone inscriptions information processing based on multi-modal knowledge graph[J]. Computers & electrical engineering, 2021, 92: 107173.
|
| [47] |
江铭虎,邓北星,廖盼盼,等.甲骨文字库与智能知识库的建立[J].计算机工程与应用,2004(4):45-47,60.
|
| [48] |
甲骨文信息处理重点实验室[EB/OL]. [2021-04-09]. http://jgwsys.aynu.edu.cn/index.htm.
|
| [49] |
熊晶,韩胜伟.甲骨文研究中跨模态知识图谱的重要性刍议[J].殷都学刊,2020,41(03):60-64,97.
|
| [50] |
Jiao Q, Jin Y, Liu Y, et al. Module structure detection of oracle characters with similar semantics[J]. Alexandria engineering journal, 2021, 60(5): 4819-4828.
|
| [51] |
顾绍通.基于分形几何的甲骨文字形识别方法[J].中文信息学报,2018,32(10):138-142.
|
| [52] |
刘志基.简论甲骨文字频的两端集中现象[J].语言研究,2010,30(4):114-122.
|
| [53] |
李邦,刘永革.文献数字化技术在甲骨文数据库建设中的应用与展望[J].殷都学刊,2020,41(3):47-53.
|
| [54] |
赵薇.作为计算批评的数字人文[J].中国文学批评,2022(2):157-166,192.
|
| [55] |
LIU A. Where is cultural criticism in the digital humanities?[M]. GOLD M K. Debates in the digital humanities. Minneapolis: University of Minnesota Press, 2012:495-501.
|
/
| 〈 |
|
〉 |