基于内容的商标图像检索
|
摘要
基于内容的图像检索(Content Based Image Retrieval，简称CBIR)，是指直接根据描述图像内容的各种特征进行图像检索的技术，目前主要集中于底层特征的相似度匹配的研究，包括颜色、纹理、形状等。CBIR技术可应用于图像数据库、电子图书馆、商标管理、公安系统等多种领域，其应用广泛使得对CBIR的研究已成为计算机视觉、图像数据库与知识挖掘等领域最活跃的研究热点之一。
本文首先概述了目前国内外基于内容的图像检索领域的现状和发展趋势，分析了现有的基于颜色、纹理、形状的图像检索方法及其特点之后，考虑到本文是对二值商标图像进行研究，所以我们选择基于形状的方法进行研究。然后本文对CBIR系统模型和常用的评价准则进行了分析，选取PVR曲线作为本文的检索性能评价标准。本文针对不同的子图像融合准则进行了实验，选取加权最小平均值准则作为本文的子图像融合准则。
为了充分利用商标图像的内部信息来提高商标图像的检索精度，本文提出了一种改进的检索方法，即融合商标子图像的多个形状特征及其内部空间关系特征来进行检索。为了证明该方法的有效性，本文分别作了基于单个形状特征(面积)、两个形状特征(面积、曲折度)融合、两个形状特征(面积、曲折度)与空间关系特征(子图像重心之间距离)融合的子图像融合实验。实验中我们选取了三个不同的商标库，商标库中分别包括3000、4000、5000多幅二值商标图像。实验结果证明，本文提出的方法即融合子图像的多个形状特征与空间关系特征的商标检索性能最好，最符合人眼视觉感受。
本文设计了一套基于内容的商标检索系统，主要是作为一个实验性的测试平台。系统的开发平台为Microsoft Windows 2000 Server，开发工具为Visual C++6.0。
Content Based Image Retrieval (CBIR) means retrieving images based on the image content description features directly. The present research on CBIR focuses on the similarity matching of lower-features, such as color, texture, shape, etc. CBIR can be applied to image database, economic library, brand management, and public security systems. It is obvious important and significant that CBIR has been one of most active researches in computer vision, image database, knowledge discovery and so on.
The present and the future of CBIR and the key issues are discussed in this paper, after analyzed the retrieval methods and characteristics of color-based, texture-based, shape-based, this paper's research aim at the method of shape-based. And after analyzed the model and common estimate rules, this paper selects PVR curve as estimate rule. This paper does some experiments to select the right sub-image integrate rules, the result of experiments show the rule of add weight to minimal-average is the best one.
This paper presents a trademark retrieval method in which the several shape features and spatial relationship are both used for the purpose of making full use of image inner info and improving retrieval precision. To prove the validity of the method, this paper does the experiments for single-shape feature based, two-shape features integration, two-shape features and spatial relationship integration. The experiments base on three trademark libraries, which respective includes more than 3000, 4000, 5000 binary trademarks. The results of experiments show the method has best precision and the output accords with people's visual feeling best.
This paper designs a Content Based Trademark Image Retrieval system as a test platform. The system platform for developing this CBIR system is Windows 2000 Server, and the development environment is Visual C++ 6.0.
本文首先概述了目前国内外基于内容的图像检索领域的现状和发展趋势，分析了现有的基于颜色、纹理、形状的图像检索方法及其特点之后，考虑到本文是对二值商标图像进行研究，所以我们选择基于形状的方法进行研究。然后本文对CBIR系统模型和常用的评价准则进行了分析，选取PVR曲线作为本文的检索性能评价标准。本文针对不同的子图像融合准则进行了实验，选取加权最小平均值准则作为本文的子图像融合准则。
为了充分利用商标图像的内部信息来提高商标图像的检索精度，本文提出了一种改进的检索方法，即融合商标子图像的多个形状特征及其内部空间关系特征来进行检索。为了证明该方法的有效性，本文分别作了基于单个形状特征(面积)、两个形状特征(面积、曲折度)融合、两个形状特征(面积、曲折度)与空间关系特征(子图像重心之间距离)融合的子图像融合实验。实验中我们选取了三个不同的商标库，商标库中分别包括3000、4000、5000多幅二值商标图像。实验结果证明，本文提出的方法即融合子图像的多个形状特征与空间关系特征的商标检索性能最好，最符合人眼视觉感受。
本文设计了一套基于内容的商标检索系统，主要是作为一个实验性的测试平台。系统的开发平台为Microsoft Windows 2000 Server，开发工具为Visual C++6.0。
Content Based Image Retrieval (CBIR) means retrieving images based on the image content description features directly. The present research on CBIR focuses on the similarity matching of lower-features, such as color, texture, shape, etc. CBIR can be applied to image database, economic library, brand management, and public security systems. It is obvious important and significant that CBIR has been one of most active researches in computer vision, image database, knowledge discovery and so on.
The present and the future of CBIR and the key issues are discussed in this paper, after analyzed the retrieval methods and characteristics of color-based, texture-based, shape-based, this paper's research aim at the method of shape-based. And after analyzed the model and common estimate rules, this paper selects PVR curve as estimate rule. This paper does some experiments to select the right sub-image integrate rules, the result of experiments show the rule of add weight to minimal-average is the best one.
This paper presents a trademark retrieval method in which the several shape features and spatial relationship are both used for the purpose of making full use of image inner info and improving retrieval precision. To prove the validity of the method, this paper does the experiments for single-shape feature based, two-shape features integration, two-shape features and spatial relationship integration. The experiments base on three trademark libraries, which respective includes more than 3000, 4000, 5000 binary trademarks. The results of experiments show the method has best precision and the output accords with people's visual feeling best.
This paper designs a Content Based Trademark Image Retrieval system as a test platform. The system platform for developing this CBIR system is Windows 2000 Server, and the development environment is Visual C++ 6.0.
引文
[1] Y. Jie, W. Alex. Beyond HCI: Multimodal manipulation of multimedia information. 计算机学报. 23(12): 1245-1252
[2] V. N. Gudivada, J. V. Raghavan. Special issue on content-based image retrieval systems. IEEE Computer Magazine 1995 28(9): 18-62
[3] Anil K. Jain, and Aditya Vailaya. Shape-based retrieval: A case study with trademark image database. Pattern Recognition. 1998 31(9): 1369-1390
[4] Y. Rui, T. S. Huang, S. Chang. Image retrieval: Current techniques, promising directions and open issues. Journal of Visual Communication and Image Representation. 1999 10:1-23
[5] Reed T. R., Buf J. M. H.. A review of recent texture segmentation and feature extraction techniques. CVGIP-IU. 199357(3): 359-372
[6] I. K. Sethi, I. Coman, B. Day, F. Jiang, D. Li, J. S. Juarez, G. Wei, B. You. Color-WISE: A system for image similarity retrieval using color. SPIE. 1997 3312: 140-149
[7] Y. J. Zhang, Z. W. Liu, Y. He. Comparison and improvement of color-based image retrieval techniques. SPIE.1997 3312:371-382
[8] Michael J. Swain, Danna H. Ballad. Color indexing. International Journal of Computer Vision, 1991 7(1): 11-32
[9] B. M. Mehtre, M. S. Kankanhalli. Color matching for image retrieval. Pattern Recognition Letters. 1995 16(3): 325-331
[10] W. Y. Ma, B. S. Manjunath. Netra: A toolbox for navigating large image databases. In Proc. IEEE Int. Conf. on Image Processing, 1997
[11] J. Harfner, H. S. Sawhney, W. Equitz, M. M. Flickner, W. Niblack. Efficient color histogram indexing for quadratic form distance functions. IEEE Trans. on Pattern Analysis and Machine Intelligence. 1995 17(7): 729-736
[12] C. Li, J. R. Smith, L. D. Bergman, V. Castelli. Sequential processing for content-based retrieval of composite objects. SPIE. 1997 3312: 2-13
[13] A. Siebert. Segmentation based image retrieval. SPIE. 1997 3312:14-24
[14] Kenneth. R. Castleman. 数字图像处理，电子工业出版社，北京，1998
[15] Yu Zhong, Anil K. Jain. Object localization using color, texture and shape. Pattern
Recognition. 2000 33:671-684
[16] W. Niblack, R. Barber, W. Equitz, M. D. Flickner, E. H. Glasman, D. Petkovic, P. Yanker, C. Faloutsos, G. Taubin, Y. Heights. Querying images by content, using color, texture, and shape. SPIE Conference on Storage and Retrieval for Image and Video Database, 1993 1908: 173-187
[17] W. Niblack, X. Zhu, J. L. Hafner, T. Breuel, D. Ponceleon, D. Petkovic, M. Flickner, E. Upfal, S. I. Nin, S. Sull, B. Dom, B. Yeo, S. Srinivasan, and M. Penner. Updates to the QBIC system. SPIE. 1997 3312:150-161
[18] S. F. Chang, A. Eleftheriadis, R. McClintock. Next-generation content representation, creation and searching for new media applications in education. IEEE Proceedings 1998 86(5): 602-615
[19] Haria BartoLini, Paolo Ciaccia etc. Feedback By Pass:A new Approach to Interactive Similarity Query Processing. In Proceedings of the 27th VLDB Conference Roma, Italy, 2001
[20] 孙兴华，郭丽，郭跃飞，杨静宇。基于目标区域的彩色图像检索研究。计算机研究与发展。2001 38(9)：1112-1120
[21] 吴翌，庄越挺，潘云鹤。视频的检索反馈。计算机研究与发展。2001 38(5)：546-551
[22] 卢汉清，孔维新，廖明，马颂德。基于内容的视频信号与图像库检索中的图像技术。自动化学报2001 27(1)：56-69
[23] 庄越挺，潘云鹤。基于内容的图像检索综述。模式识别与人工智能。1999 12(2)：170-177
[24] 章毓晋。图像处理和分析。北京：清华大学出版社，1999．
[25] 章毓晋。图像理解与计算机视觉。北京：电子工业出版社，2000
[26] 徐旭。基于视觉特征的图像检索系统研究[博士学位论文]。浙江大学，杭州，1999．
[27] 柏正尧。基于矩方法的图像检索研究[硕士学位论文]。云南大学，云南，2000
[28] 章毓晋。图像工程(上册)—图像处理和分析。北京：清华大学出版社，1999
[29] 章毓晋。图像分割。北京：科学出版社，2001
[30] 马跃先，胡晓改，秦晋渊。二值商标图像检索系统。山西大学学报(自然科学版)。1998 21(2)：137-143
[31] 乃学武。基于内容图像检索中的关键技术的研究[硕士学位论文]。西安电子科技大学，西安 2002
[32] 任明武。复杂环境图像处理的若干算法研究及船舶目标检测系统[博士学位论文]。南京理工大学，南京 2001
[33] 黄元元，郭丽，杨靖宇。基于形状与空间特征的二值商标图像检索。模式识别与人工智能。2003 16(2)：241-245
[34] 刘治锋。远程水情监控的水位识别和手写体数字识别的特征提取[硕士学位论文]。南京理工大学，南京2004
[35] 罗希平，田捷，诸葛婴，王靖，戴汝为。图像分割方法综述。模式识别与人工智能。1999 12(3)：300-312
[36] 孙兴华。基于内容的图像检索研究[博士论文]。南京理工大学。南京 2001
[37] 刘忠伟，章毓晋。综合利用颜色和纹理特征的图像检索。通信学报。1999 20(5)：36-40
[38] 周丽华，周纪勤，柏正尧。基于图像信息熵的商标图像检索。计算机应用。200020(2)：21-23
[2] V. N. Gudivada, J. V. Raghavan. Special issue on content-based image retrieval systems. IEEE Computer Magazine 1995 28(9): 18-62
[3] Anil K. Jain, and Aditya Vailaya. Shape-based retrieval: A case study with trademark image database. Pattern Recognition. 1998 31(9): 1369-1390
[4] Y. Rui, T. S. Huang, S. Chang. Image retrieval: Current techniques, promising directions and open issues. Journal of Visual Communication and Image Representation. 1999 10:1-23
[5] Reed T. R., Buf J. M. H.. A review of recent texture segmentation and feature extraction techniques. CVGIP-IU. 199357(3): 359-372
[6] I. K. Sethi, I. Coman, B. Day, F. Jiang, D. Li, J. S. Juarez, G. Wei, B. You. Color-WISE: A system for image similarity retrieval using color. SPIE. 1997 3312: 140-149
[7] Y. J. Zhang, Z. W. Liu, Y. He. Comparison and improvement of color-based image retrieval techniques. SPIE.1997 3312:371-382
[8] Michael J. Swain, Danna H. Ballad. Color indexing. International Journal of Computer Vision, 1991 7(1): 11-32
[9] B. M. Mehtre, M. S. Kankanhalli. Color matching for image retrieval. Pattern Recognition Letters. 1995 16(3): 325-331
[10] W. Y. Ma, B. S. Manjunath. Netra: A toolbox for navigating large image databases. In Proc. IEEE Int. Conf. on Image Processing, 1997
[11] J. Harfner, H. S. Sawhney, W. Equitz, M. M. Flickner, W. Niblack. Efficient color histogram indexing for quadratic form distance functions. IEEE Trans. on Pattern Analysis and Machine Intelligence. 1995 17(7): 729-736
[12] C. Li, J. R. Smith, L. D. Bergman, V. Castelli. Sequential processing for content-based retrieval of composite objects. SPIE. 1997 3312: 2-13
[13] A. Siebert. Segmentation based image retrieval. SPIE. 1997 3312:14-24
[14] Kenneth. R. Castleman. 数字图像处理，电子工业出版社，北京，1998
[15] Yu Zhong, Anil K. Jain. Object localization using color, texture and shape. Pattern
Recognition. 2000 33:671-684
[16] W. Niblack, R. Barber, W. Equitz, M. D. Flickner, E. H. Glasman, D. Petkovic, P. Yanker, C. Faloutsos, G. Taubin, Y. Heights. Querying images by content, using color, texture, and shape. SPIE Conference on Storage and Retrieval for Image and Video Database, 1993 1908: 173-187
[17] W. Niblack, X. Zhu, J. L. Hafner, T. Breuel, D. Ponceleon, D. Petkovic, M. Flickner, E. Upfal, S. I. Nin, S. Sull, B. Dom, B. Yeo, S. Srinivasan, and M. Penner. Updates to the QBIC system. SPIE. 1997 3312:150-161
[18] S. F. Chang, A. Eleftheriadis, R. McClintock. Next-generation content representation, creation and searching for new media applications in education. IEEE Proceedings 1998 86(5): 602-615
[19] Haria BartoLini, Paolo Ciaccia etc. Feedback By Pass:A new Approach to Interactive Similarity Query Processing. In Proceedings of the 27th VLDB Conference Roma, Italy, 2001
[20] 孙兴华，郭丽，郭跃飞，杨静宇。基于目标区域的彩色图像检索研究。计算机研究与发展。2001 38(9)：1112-1120
[21] 吴翌，庄越挺，潘云鹤。视频的检索反馈。计算机研究与发展。2001 38(5)：546-551
[22] 卢汉清，孔维新，廖明，马颂德。基于内容的视频信号与图像库检索中的图像技术。自动化学报2001 27(1)：56-69
[23] 庄越挺，潘云鹤。基于内容的图像检索综述。模式识别与人工智能。1999 12(2)：170-177
[24] 章毓晋。图像处理和分析。北京：清华大学出版社，1999．
[25] 章毓晋。图像理解与计算机视觉。北京：电子工业出版社，2000
[26] 徐旭。基于视觉特征的图像检索系统研究[博士学位论文]。浙江大学，杭州，1999．
[27] 柏正尧。基于矩方法的图像检索研究[硕士学位论文]。云南大学，云南，2000
[28] 章毓晋。图像工程(上册)—图像处理和分析。北京：清华大学出版社，1999
[29] 章毓晋。图像分割。北京：科学出版社，2001
[30] 马跃先，胡晓改，秦晋渊。二值商标图像检索系统。山西大学学报(自然科学版)。1998 21(2)：137-143
[31] 乃学武。基于内容图像检索中的关键技术的研究[硕士学位论文]。西安电子科技大学，西安 2002
[32] 任明武。复杂环境图像处理的若干算法研究及船舶目标检测系统[博士学位论文]。南京理工大学，南京 2001
[33] 黄元元，郭丽，杨靖宇。基于形状与空间特征的二值商标图像检索。模式识别与人工智能。2003 16(2)：241-245
[34] 刘治锋。远程水情监控的水位识别和手写体数字识别的特征提取[硕士学位论文]。南京理工大学，南京2004
[35] 罗希平，田捷，诸葛婴，王靖，戴汝为。图像分割方法综述。模式识别与人工智能。1999 12(3)：300-312
[36] 孙兴华。基于内容的图像检索研究[博士论文]。南京理工大学。南京 2001
[37] 刘忠伟，章毓晋。综合利用颜色和纹理特征的图像检索。通信学报。1999 20(5)：36-40
[38] 周丽华，周纪勤，柏正尧。基于图像信息熵的商标图像检索。计算机应用。200020(2)：21-23
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |