Please wait a minute...
J4  2012, Vol. 46 Issue (3): 392-401    DOI: 10.3785/j.issn.1008-973X.2012.03.003
计算机技术     
基于构图分析的古代壁画相关度评价方法
王琦, 鲁东明
浙江大学 计算机科学与技术学院,浙江 杭州 310027
Composition analysis-based relevance ranking for ancient mural
WANG Qi, LU Dong-ming
College of Computer Science and Technology, Zhejiang University, Hangzhou 310027, China
 全文: PDF  HTML
摘要:

由于目前的图像检索技术没有考虑壁画的构图学特征,缺乏对复杂语义的处理能力,难以满足古代壁画研究工作对检索全面性和准确性的要求.为提高古代壁画图像语义检索的质量,提出基于构图分析的相关度模型,通过引入基于绘画构图学的理论和分析方法,从壁画内容的布局、主题和语义三方面用量化方法描述检索语义与壁画内容的相关度,较好地解决了用户的真实检索意图与壁画内容间的“语义鸿沟”问题.该相关度评价模型可嵌入基于语义查询扩展的框架中,以提高Top N结果的准确率,同时维持了较高的查全率.敦煌壁画资料检索的实际应用表明:以反映前n个结果准确率的R-Precision为评测指标,基于构图分析的相关度评价方法可比未采用相关度评价的基线方法平均高出36%.

Abstract:

The present image retrieval technologies have difficulties in retrieving ancient murals, since they lack of the abilities to handle complex semantic and features of layout in painting. This work puts forward a new relevance ranking model based on composition analysis to improve ancient mural retrieval. By introducing the theory of composition on painting, the relevance ranking model measures the relevance of mural images from three aspects which are layout, topic and semantics, and reduces the semantic gap between the content of mural and the real intention of the user. The relevance ranking model was seamlessly integrated into a unified framework for semantic query expansion to improve the precision of Top N results while maintaining a high recall. Experimental results of the Dunhuang Murals show that compared with the baseline method, the R-Precision ratio of semantic mural retrieval based on this model can be increased by 36% on average.

出版日期: 2012-03-01
:  TP 391  
基金资助:

国家教育部长江学者和创新团队发展计划资助项目(IRT0652);教育科研基础设施IPv6技术升级和应用示范项目;古代壁画保护国家文物局重点科研基地开放课题资助项目;新世纪优秀人才支持计划资助项目(NCET- 04-0535).

通讯作者: 鲁东明,男,教授,博导.     E-mail: ldm@cs.zju.edu.cn
作者简介: 王琦(1974-),男,教授级高级工程师,从事数字化文物与艺术研究. E-mail: wangchee99@yahoo.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  

引用本文:

王琦, 鲁东明. 基于构图分析的古代壁画相关度评价方法[J]. J4, 2012, 46(3): 392-401.

WANG Qi, LU Dong-ming. Composition analysis-based relevance ranking for ancient mural. J4, 2012, 46(3): 392-401.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2012.03.003        http://www.zjujournals.com/eng/CN/Y2012/V46/I3/392

[1] 王崇骏,杨育彬,陈世福. 基于高层语义的图像检索算法 [J]. 软件学报,2004,15(10):1461-1469.
WANG Chongjun, YANG Yubin, CHEN Shifu. Algorithms of highlevel semanticbased image retrieval [J]. Journal of Software, 2004, 15(10): 1461-1469.
[2] JIANG Shuqiang, HUANG Tiejun, GAO Wen. An ontologybased approach to retrieve digitized art images[C]∥ Proceedings of the Web Intelligence (IEEE/WIC/ACM WI 04). Washington, DC, USA: IEEE Computer Society, 2004: 131-137.
[3]  JIANG Shuqiang, DU Jun, HUANG Qingming, et al. Visual ontology construction for digitized art image retrieval [J]. Journal of Computing Science and Technology, 2005, 20(6): 855-860.
[4] DENG J, DONG W, SOCHER R, et al. ImageNet: a largescale hierarchical image database [C]∥ Proceedings of the Computer Vision and Pattern Recognition (CVPR) 2009. Washington, DC, USA: IEEE Computer Society, 2009: 248-255.
[5] LI LJ, SOCHER R AND LI FF. Towards total scene understanding: classification, annotation and segmentation in an automatic framework [C]∥ Proceedings of the Computer Vision and Pattern Recognition (CVPR) 2009. Washington, DC, USA: IEEE Computer Society, 2009: 2036-2043.
[6] SOCHER R AND LI FF. Connecting modalities: semisupervised segmentation and annotation of images using unaligned text corpora [C]∥ Proceedings of the IEEE Computer Vision and Pattern Recognition (CVPR) 2010. Washington, DC, USA: IEEE Computer Society, 2010: 966-973.
[7]  NATSEV A, HAUBOLD A, TEI J, et al. Semantic conceptbased query expansion and reranking for multimedia retrieval [C]∥ Proceedings of the 15th international conference on Multimedia. New York, NY, USA: ACM, 2007: 25-29.
[8]  张鸿,吴飞,庄越挺,等.一种基于内容相关性的跨媒体检索方法 [J].计算机学报,2008,31(5):820-826.
ZHANG Hong, WU Fei, ZHUANG Yueting, et al. Crossmedia retrieval method based on content correlations [J]. Chinese Journal of Computers, 2008, 31(5): 820-826.
[9]  王梅,周向东,张军旗,等.基于扩展生成语言模型的图像自动标注方法 [J].软件学报, 2008, 19(9) : 2449-2460.
WANG Mei, ZHOU Xiangdong, ZHANG Junqi, et al. Image autoannotation via an extended generative language method [J]. Journal of Software, 2008, 19(9): 2449-2460.
[10]  蒋跃. 绘画构图学教程[M].杭州:中国美术学院出版社,2003:1.
[11]  赵声良. 敦煌艺术十讲[M].上海:上海古籍出版社,2007:179.
[12]  BAEZAYATES R, RIBEIRONETO B. Modern information retrieval [M].NewYork:AddisonWesleyLongman,1999:73-97.
[13] CUTLER M, SHIH Y, MENG W. Using the structure of HTML documents to improve retrieval [C]∥ Proceedings of the USENIX Symposium on Internet Technologies and Systems (NISTS’97). California, USA: \
[s.n.\], 1997: 241-251.
[14]  韩玮.中国画构图艺术[M].山东:山东美术出版社,2002:87-88.
[15]  史敦宇,金洵瑨. 敦煌舞乐线描集[M]. 甘肃:甘肃人民美术出版社, 2007:117.
[16]  MATSUO Y, ISHIZUKA M. Keyword extraction from a single document using word cooccurrence statistical information[C]∥ Proceedings of the Sixteenth International Florida Artificial Intelligence Research Society Conference. California, USA: AAAI, 2003:392-396.
[17]  DOUG Beeferman, ADAM Berger, JOHN Lafferty. A model of lexical attraction and repulsion [C]∥ Proceedings of the 35th Annual Meeting of the Association for Computational Linguistics. Madrid, Spain:\
[s.n.\], 1997:373-380.
[18] LEACOCK C, CHODOROW M. WordNet: A Lexical reference system and its application [M]. London, England: MIT Press, 1998: 265-283.
[19] HOLLINK L, SCHREIBER G, WIELEMAKER J, et al. Semantic annotation of image collections [C] ∥Proceedings of the KCAP'03 Workshop on Knowledge Capture and Semantic Annotation. Sanibel, Florida, USA: ACM, 2003: 41-48.
[20] 胡同庆.敦煌石窟艺术概述[J].敦煌研究,1993(3):16-34.
HU Tongqing. An introduction to Dunhuang Grotto art [J]. Dunhuang Research, 1993 (3): 16-34.
[21]  敦煌文物研究所.中国石窟敦煌莫高窟[M].北京:文物出版社,平凡社, 1987.
[22]  酒井敦子.南北朝时期的植物云气纹样 [J].敦煌研究, 2003 (2):20-28.
SAKAI Atsko. The motifs of scrolling flora in the Northern and the Southern Dynasties [J]. Dunhuang Research, 2003(2): 20-28.

[1] 赵建军,王毅,杨利斌. 基于时间序列预测的威胁估计方法[J]. J4, 2014, 48(3): 398-403.
[2] 张天煜, 冯华君, 徐之海, 李奇, 陈跃庭. 基于强边缘宽度直方图的图像清晰度指标[J]. J4, 2014, 48(2): 312-320.
[3] 刘中, 陈伟海, 吴星明, 邹宇华, 王建华. 基于双目视觉的显著性区域检测[J]. J4, 2014, 48(2): 354-359.
[4] 崔光茫, 赵巨峰, 冯华君, 徐之海, 李奇, 陈跃庭. 非均匀介质退化图像快速仿真模型的建立[J]. J4, 2014, 48(2): 303-311.
[5] 王相兵,童水光,钟崴,张健. 基于可拓重用的液压挖掘机结构性能方案设计[J]. J4, 2013, 47(11): 1992-2002.
[6] 王进, 陆国栋, 张云龙. 基于数量化一类分析的IGA算法及应用[J]. J4, 2013, 47(10): 1697-1704.
[7] 刘羽, 王国瑾. 以已知曲线为渐进线的可展曲面束的设计[J]. J4, 2013, 47(7): 1246-1252.
[8] 胡根生,鲍文霞,梁栋,张为. 基于SVR和贝叶斯方法的全色与多光谱图像融合[J]. J4, 2013, 47(7): 1258-1266.
[9] 吴金亮, 黄海斌, 刘利刚. 保持纹理细节的无缝图像合成[J]. J4, 2013, 47(6): 951-956.
[10] 陈潇红,王维东. 基于时空联合滤波的高清视频降噪算法[J]. J4, 2013, 47(5): 853-859.
[11] 朱凡,李悦,蒋 凯,叶树明,郑筱祥. 基于偏最小二乘的大鼠初级运动皮层解码[J]. J4, 2013, 47(5): 901-905.
[12] 吴宁, 陈秋晓, 周玲, 万丽. 遥感影像矢量化图形的多层次优化方法[J]. J4, 2013, 47(4): 581-587.
[13] 计瑜,沈继忠,施锦河. 一种基于盲源分离的眼电伪迹自动去除方法[J]. J4, 2013, 47(3): 415-421.
[14] 王翔,丁勇. 基于Gabor滤波器的全参考图像质量评价方法[J]. J4, 2013, 47(3): 422-430.
[15] 刘芳, 孙芸, 杨庚, 林海. 基于粒子群优化算法的社交网络可视化[J]. J4, 2013, 47(1): 37-43.