基于空间关系的图像检索方法研究
|
摘要
随着计算机技术的广泛应用以及计算机技术与其它领域相关技术的充分结合，在新闻出版、医疗卫生、建筑设计等各个行业每天会有产生大量的数字图像。这些图像是一种宝贵的资源，如何从这一丰富的资源中获取感兴趣的信息是一个必须解决的问题。这就是图像检索问题。传统的基于字符的精确匹配检索方法并不能很好的应用于图像检索。因此，基于图像内容的检索方法研究成为研究的热点。本文将基于内容的图像检索方法作为研究对象，在以下方面进行了一些研究性的工作。
1、将图像的空间关系用拓扑关系、投影关系、角度关系来表示，并给出了以上几种关系的定义。
2、分别给出了拓扑关系、投影关系、角度关系距离的定义。
3、分析了平移、放缩、旋转变换对基于对象空间关系图像检索方法的影响。
4、通过求查询图像与数据库图像中共有对象数目的均值，将查询图像与数据库图像中共有对象数目加入到图像相似程度的度量尺度中。
5、给出了两幅图像相似程度的度量公式及检索算法。
6、将权重引入到相似程度度量公式当中，通过权重调整，可实现满足用户某方面需求的图像检索。
7、开发了一个基于空间关系的图像检索系统，对图像的检索结果进行了分析，并与其它的检索方法进行了比较。
基于内容的图像检索作为一个较新的研究领域，检索方法都不很成熟。如何使图像检索系统与人类判别图像相似程度的方法相一致，既合理选择表征图像的特征还需进一步研究。
As technologies of qomputer are widely applied, and technologies of computer combine with other technology of field relating, the fields of publishing industry, medical and related health professions and architectural and engineering design produce lots of digital image respectively every day. These images are precious resources. How to make fully use of these plenty of resources and how to easily get information we are interested in is a must solution. It is a question of image retrieval. The classical method of text-based retrieval is not suitable for image retrieval. Therefore, method of content-based image retrieval is widely studied. In this thesis, we present a method of image retrieval. We study from several aspects, which include:
1. We represent image with topology relation, project relation
and angle relation. And we present their definition respectively.
2. We present distance definition between each spatial relation
respectively.
3. The effect of translation, scale and rotation is analyzed.
4. The number of common object between query image and
database image is considered.
5. A formulation counting similar degree between two images is
presented. A retrieval algorithm based on spatial relations among image objects is implemented.
6. Weight is added to retrieval algorithm. All kinds of users are
satisfied with query result through adjusting weight of spatial relation.
7. A system of image retrieval based on spatial relation is
implemented. A comparison with other method is presented
too.
Content-based image retrieval is a very new research field; therefore its retrieval method is not mature. How to make the image retrieval system works as accurately as mankind do (namely reasonably selecting the characteristics of the image) has much room to be studied further.
1、将图像的空间关系用拓扑关系、投影关系、角度关系来表示，并给出了以上几种关系的定义。
2、分别给出了拓扑关系、投影关系、角度关系距离的定义。
3、分析了平移、放缩、旋转变换对基于对象空间关系图像检索方法的影响。
4、通过求查询图像与数据库图像中共有对象数目的均值，将查询图像与数据库图像中共有对象数目加入到图像相似程度的度量尺度中。
5、给出了两幅图像相似程度的度量公式及检索算法。
6、将权重引入到相似程度度量公式当中，通过权重调整，可实现满足用户某方面需求的图像检索。
7、开发了一个基于空间关系的图像检索系统，对图像的检索结果进行了分析，并与其它的检索方法进行了比较。
基于内容的图像检索作为一个较新的研究领域，检索方法都不很成熟。如何使图像检索系统与人类判别图像相似程度的方法相一致，既合理选择表征图像的特征还需进一步研究。
As technologies of qomputer are widely applied, and technologies of computer combine with other technology of field relating, the fields of publishing industry, medical and related health professions and architectural and engineering design produce lots of digital image respectively every day. These images are precious resources. How to make fully use of these plenty of resources and how to easily get information we are interested in is a must solution. It is a question of image retrieval. The classical method of text-based retrieval is not suitable for image retrieval. Therefore, method of content-based image retrieval is widely studied. In this thesis, we present a method of image retrieval. We study from several aspects, which include:
1. We represent image with topology relation, project relation
and angle relation. And we present their definition respectively.
2. We present distance definition between each spatial relation
respectively.
3. The effect of translation, scale and rotation is analyzed.
4. The number of common object between query image and
database image is considered.
5. A formulation counting similar degree between two images is
presented. A retrieval algorithm based on spatial relations among image objects is implemented.
6. Weight is added to retrieval algorithm. All kinds of users are
satisfied with query result through adjusting weight of spatial relation.
7. A system of image retrieval based on spatial relation is
implemented. A comparison with other method is presented
too.
Content-based image retrieval is a very new research field; therefore its retrieval method is not mature. How to make the image retrieval system works as accurately as mankind do (namely reasonably selecting the characteristics of the image) has much room to be studied further.
引文
[1] Niblack w. Barber R. et al. The QBICProject: Querying Images by Content Using Color. Texture and Shape. In: Proc. SPIE Conf on Visual Commun and Image Proc, 1994, 203-207
[2] Bach J R. Fuller C. Gupta A, et al. The Virage Image Search Engine: An Open Framework for Image Management. Pro SPIE Conf on Visual Commun and Image Proc. 1995 400-406
[3] John Eakins. Margaret Graham Content-based Image Retrieval Report99 1999． 10 22-26
[4] Swain M J.Ballard D H. Color indexing. Intl. J. Computer Vision. 1991 (7) ,11-32
[5] Strieker M. Orengo M. Similarity of color image. SPIE. 1995． 2420:381-392．
[6] W. Niblack. The QBIC project: Querying images by content using color. texture, and shape. SPIE, 1993; V1908:173-187
[7] Hafner J, Sawhney H S, Equitz W, Flicker M, Niblack W. Efficent Color Histogram Indexing for Quadratic Form Distance Functions. IEEE Trans on PAMI, 1995． (17) ,729-736
[8] Haralick R M, Shanmugam K, Dinstein I. Textural Features for Image Classification.IEEE Trans on SMC,1973,3(6) ,768-780
[9] Tamura, H et al Textural features corresponding to visual perception IEEE Trans. On SMC 8(6) 460-472
[10] Kashyap R L, Khotanzad A. A Model-based Method for Rotation Invariant Texure Classification. IEEE Trans. Pattern Analysis and Machine Intelligence 1986,8,472-481
[11] Cohen F S, Fan Z, Patel M. Classification of Rotated and Scaled Textured images Using Gaussian Markov Random Field Models. IEEE Trans. Pattern Analysis and Machine Intelligence, 1991,13,192-202
[12] Panjwani D K, Healey G. Markov Random Field Models for Unsupervised Segmentation of Textured Color. IEEE Trans. Pattern Analysis and Machine Intelligence, 1995,17(10)
[13] Pentland, Fractal-based description of natural scenes, IEEE Trans. Pattern Analysis and Machine Intelligence. 6(6) 661-674, November 1984
[14] Gross M II, Koch R, Lippert L. Dreger A. Multiscale Image Texture Analysis in Wavelet Spaces . Proc IEEE Intl Conf on Image Proc. 1994, 168-174
[15] B.S. Manjunath W.Y. Ma Texture Features for Browsing and Retrieval of Image Data IEEE Trans. Pattern Analysis and Machine Intelligence. 1 8(8)
1996，837-842
[16]贾云得 编著 机器视觉 科学出版社 2000．4
[17]王润生 编著 图像理解 国防科技大学出版社 1995．10
[18]Barrow Ⅱ G. Parametric Correspondence and Chamfer Matching: Two New Techniques for Image Matching. Proc 5th Int Joint Conf Artificial, 1977, 134-139
[19]Borgefors G. Hierarchical Chamfer Mathcing: A Parametric Edge Matching Algorithm. IEEE Trans Patt Recog and Mach Intell, 1988, 16(8)32-50
[20]邬浩、潘云鹤、庄越挺、杨宇艇 基于对象形状的图象查询技术 软件学报 1998，9(5)343-349
[21]H. Tamura and N.Yokoya, Image database systems a survey, Pattern Recognition 17, 1984, 29-43
[22]S.L. Tanimoto. An Iconic/Symbolic Data Structuring Scheme. Pattern Recognition and Artificial Intelligence, New York: Academic. 1976 452-471
[23]S.K. Chang, Q.y. Shi and C.W. Yan Iconic Indexing by 2D String. IEEE Trans. Pattern Analysis and Machine Intelligence, 9(3) 1987 413-428
[24]E. Jungert, Extended symbolic projection used in a knowledge structure for spatial reasoning, 4th BPRA Conf. On Pattern Recognition. Springer, Cambridge 1998.3
[25]S.K. Chang, E. Jungert and Y. Li, Representation and retrieval of symbolic pictures using generalized 2D-Strings, SPIE Proc. Visual Communications and Image Processing, Philadelphia, 5-10 November, 1360-1372(1989)
[26]SUH-YIN LEE ,Fang-Jung Hsu 2D-Cstring: A New Spatial Knowledge Representation for Image Database System. Pattern Recognition 23(10) 1990 1077-1087
[27]Allen J F. Maintaining knowledge about temporal interval.Communications ACM, 1983, 26 (11):832～843
[28]P.W. Huang, Y.R. Jean Using 2D-C+String as Spatial Knowledge Representation for Image Database System. Pattern Recognition 29(9) 1994 1249-1257
[29]L ee S Y, Yang M C, Chen J W. 2 D B-String:A spatial knowledge representation for image database system. In:Procof ICSC' 92 , Second International Computer ScienceConference. 1992. 609～615
[30]Mohammad Nabil, Anne H.H. Ngu, and John Shepherd Picture Similarity Retrieval Using the 2D Projection Interval Representation. IEEE Transaction on Knowledge and Data Engineering 8(4) 1996 533-539
[31]蔡少华 翟战强 GIS基础空间关系分析 测绘工程 8(2)1999．6 38-42
[32]邬伦等著 地理信息系统——原理、方法和应用 北京：科学出版社 2001．2
[33]M.J. Egenhofer and D.R. Franzosa Point-Set Topological Spatial Relations. In:Int. J. GIS, 1991, 5(2):161-174
[34]曹菡 拓扑关系渐变的定性推理 陕西师范大学学报(自然科学版)27(3)1999．9 28-32
[35]M.J. Egenhofer and K.K Al-Taha. Reasioning aboutgradual changes of topological relationships. In Proceedings of the International Conference GIS- From Space to Territory: Theories and Methods of Spatio-Tempporal Reasoning in Geographic Space. A.U. Frank. I. Campari, and U. Formentini. Berlin:Springer- Verlag, 1992. 196-219.
[36]J. F. Allen Maintaining Knowledge About Temporal Intervals Comm. ACM., vol. 26, no. 11. 1983 832-843
[37]舒红 博士论文 概念、形式化和逻辑时空数据建模原理初探 1998．9
[38]张师超 基于间断区间的时态知识表示 软件学报 5(6)1994 13-18
[39]C. Freksa, Temporal Reasoning Based on Semi-Intervals. Artificial Intelligence, 54(1,2) 1992 199-227
[40]PEUQUET D J, ZHAN Ci-Xiang. An Algorithmto Determine the Directional Relationship between Arbitrarily-shaped Polygons in the Plane[J]. Pat-tern Recognition, 1987, 20(1):65-74.
[41]FRANK ANDREW U. Qualitative Spatial Reason-ing About Distances and Directions in GeographicSpace[J]. J of Visual Languages and Computing, 1992,3(4):343-371
[42]FRANK ANDREW U. Qualitative Spatial Reason-ing:Cardinal Directions as An Example[J]. INT JGeographical Information Systems, 1996, 10(3):269-290
[43]曹菡 陈军 方向关系与距离关系的定性描述与推理 西安石油学院学报(自然科学版)16(1)2001．1 68-73
[44]曹菡，陈军，杜道生 空间目标方向关系的定性扩展描述 测绘学报30(2) 2001．5 163-167
[45]刘忠伟 章毓晋 综合利用颜色和纹理特征的图象检索 通信学报 1999 20(5)36-40
[46]王家耀著 空间信息系统原理 北京：科学出版社 2001．5
[47]Satyaraj Pantham著 黄晓鸣等译 JFC 2D图形图像编程 北京：电子工业出版社 2000．07
[48]严蔚敏著 数据结构 北京：清华大学出版社 1991．10
[49]Smith JR, Chang S F. Tools and techniques for color image retrieval. In: Proc of SPIE: Storage and Retrieval for Image and Video Database. San Jose, CA, 1996, 26(70)426-437
[50]Tversky A. Feature of similarity. Psychological Review, 1977, 84(4): 327-352
[51]倪明田 计算机图形学 北京：北京大学出版社 1999．11
[52]Essam A. El-kwae A Robust Framework for Content-Based Retrieval by Spatial Similarity in Image Database. Acm Transactions on Information Systems 1999.4 17(2) 174-198
[53]易江涛 硕士学位论文 基于内容的图象检索系统的设计与实现 中国科学技术大学 1999．4
[54]张海藩著 软件工程导论 北京：清华大学出版社 1999．5
[55]黄上腾 数据库原理 上海：上海交通大学出版社 1996．8
[56]Clifford A. Shaffer著 张铭等译 数据结构与算法分析(java版) 北京：电子工业出版社 2001．2
[57]Bruce Eckel著 京京工作室译 Java编程思想 北京：机械工业出版社 1999．4
[2] Bach J R. Fuller C. Gupta A, et al. The Virage Image Search Engine: An Open Framework for Image Management. Pro SPIE Conf on Visual Commun and Image Proc. 1995 400-406
[3] John Eakins. Margaret Graham Content-based Image Retrieval Report99 1999． 10 22-26
[4] Swain M J.Ballard D H. Color indexing. Intl. J. Computer Vision. 1991 (7) ,11-32
[5] Strieker M. Orengo M. Similarity of color image. SPIE. 1995． 2420:381-392．
[6] W. Niblack. The QBIC project: Querying images by content using color. texture, and shape. SPIE, 1993; V1908:173-187
[7] Hafner J, Sawhney H S, Equitz W, Flicker M, Niblack W. Efficent Color Histogram Indexing for Quadratic Form Distance Functions. IEEE Trans on PAMI, 1995． (17) ,729-736
[8] Haralick R M, Shanmugam K, Dinstein I. Textural Features for Image Classification.IEEE Trans on SMC,1973,3(6) ,768-780
[9] Tamura, H et al Textural features corresponding to visual perception IEEE Trans. On SMC 8(6) 460-472
[10] Kashyap R L, Khotanzad A. A Model-based Method for Rotation Invariant Texure Classification. IEEE Trans. Pattern Analysis and Machine Intelligence 1986,8,472-481
[11] Cohen F S, Fan Z, Patel M. Classification of Rotated and Scaled Textured images Using Gaussian Markov Random Field Models. IEEE Trans. Pattern Analysis and Machine Intelligence, 1991,13,192-202
[12] Panjwani D K, Healey G. Markov Random Field Models for Unsupervised Segmentation of Textured Color. IEEE Trans. Pattern Analysis and Machine Intelligence, 1995,17(10)
[13] Pentland, Fractal-based description of natural scenes, IEEE Trans. Pattern Analysis and Machine Intelligence. 6(6) 661-674, November 1984
[14] Gross M II, Koch R, Lippert L. Dreger A. Multiscale Image Texture Analysis in Wavelet Spaces . Proc IEEE Intl Conf on Image Proc. 1994, 168-174
[15] B.S. Manjunath W.Y. Ma Texture Features for Browsing and Retrieval of Image Data IEEE Trans. Pattern Analysis and Machine Intelligence. 1 8(8)
1996，837-842
[16]贾云得 编著 机器视觉 科学出版社 2000．4
[17]王润生 编著 图像理解 国防科技大学出版社 1995．10
[18]Barrow Ⅱ G. Parametric Correspondence and Chamfer Matching: Two New Techniques for Image Matching. Proc 5th Int Joint Conf Artificial, 1977, 134-139
[19]Borgefors G. Hierarchical Chamfer Mathcing: A Parametric Edge Matching Algorithm. IEEE Trans Patt Recog and Mach Intell, 1988, 16(8)32-50
[20]邬浩、潘云鹤、庄越挺、杨宇艇 基于对象形状的图象查询技术 软件学报 1998，9(5)343-349
[21]H. Tamura and N.Yokoya, Image database systems a survey, Pattern Recognition 17, 1984, 29-43
[22]S.L. Tanimoto. An Iconic/Symbolic Data Structuring Scheme. Pattern Recognition and Artificial Intelligence, New York: Academic. 1976 452-471
[23]S.K. Chang, Q.y. Shi and C.W. Yan Iconic Indexing by 2D String. IEEE Trans. Pattern Analysis and Machine Intelligence, 9(3) 1987 413-428
[24]E. Jungert, Extended symbolic projection used in a knowledge structure for spatial reasoning, 4th BPRA Conf. On Pattern Recognition. Springer, Cambridge 1998.3
[25]S.K. Chang, E. Jungert and Y. Li, Representation and retrieval of symbolic pictures using generalized 2D-Strings, SPIE Proc. Visual Communications and Image Processing, Philadelphia, 5-10 November, 1360-1372(1989)
[26]SUH-YIN LEE ,Fang-Jung Hsu 2D-Cstring: A New Spatial Knowledge Representation for Image Database System. Pattern Recognition 23(10) 1990 1077-1087
[27]Allen J F. Maintaining knowledge about temporal interval.Communications ACM, 1983, 26 (11):832～843
[28]P.W. Huang, Y.R. Jean Using 2D-C+String as Spatial Knowledge Representation for Image Database System. Pattern Recognition 29(9) 1994 1249-1257
[29]L ee S Y, Yang M C, Chen J W. 2 D B-String:A spatial knowledge representation for image database system. In:Procof ICSC' 92 , Second International Computer ScienceConference. 1992. 609～615
[30]Mohammad Nabil, Anne H.H. Ngu, and John Shepherd Picture Similarity Retrieval Using the 2D Projection Interval Representation. IEEE Transaction on Knowledge and Data Engineering 8(4) 1996 533-539
[31]蔡少华 翟战强 GIS基础空间关系分析 测绘工程 8(2)1999．6 38-42
[32]邬伦等著 地理信息系统——原理、方法和应用 北京：科学出版社 2001．2
[33]M.J. Egenhofer and D.R. Franzosa Point-Set Topological Spatial Relations. In:Int. J. GIS, 1991, 5(2):161-174
[34]曹菡 拓扑关系渐变的定性推理 陕西师范大学学报(自然科学版)27(3)1999．9 28-32
[35]M.J. Egenhofer and K.K Al-Taha. Reasioning aboutgradual changes of topological relationships. In Proceedings of the International Conference GIS- From Space to Territory: Theories and Methods of Spatio-Tempporal Reasoning in Geographic Space. A.U. Frank. I. Campari, and U. Formentini. Berlin:Springer- Verlag, 1992. 196-219.
[36]J. F. Allen Maintaining Knowledge About Temporal Intervals Comm. ACM., vol. 26, no. 11. 1983 832-843
[37]舒红 博士论文 概念、形式化和逻辑时空数据建模原理初探 1998．9
[38]张师超 基于间断区间的时态知识表示 软件学报 5(6)1994 13-18
[39]C. Freksa, Temporal Reasoning Based on Semi-Intervals. Artificial Intelligence, 54(1,2) 1992 199-227
[40]PEUQUET D J, ZHAN Ci-Xiang. An Algorithmto Determine the Directional Relationship between Arbitrarily-shaped Polygons in the Plane[J]. Pat-tern Recognition, 1987, 20(1):65-74.
[41]FRANK ANDREW U. Qualitative Spatial Reason-ing About Distances and Directions in GeographicSpace[J]. J of Visual Languages and Computing, 1992,3(4):343-371
[42]FRANK ANDREW U. Qualitative Spatial Reason-ing:Cardinal Directions as An Example[J]. INT JGeographical Information Systems, 1996, 10(3):269-290
[43]曹菡 陈军 方向关系与距离关系的定性描述与推理 西安石油学院学报(自然科学版)16(1)2001．1 68-73
[44]曹菡，陈军，杜道生 空间目标方向关系的定性扩展描述 测绘学报30(2) 2001．5 163-167
[45]刘忠伟 章毓晋 综合利用颜色和纹理特征的图象检索 通信学报 1999 20(5)36-40
[46]王家耀著 空间信息系统原理 北京：科学出版社 2001．5
[47]Satyaraj Pantham著 黄晓鸣等译 JFC 2D图形图像编程 北京：电子工业出版社 2000．07
[48]严蔚敏著 数据结构 北京：清华大学出版社 1991．10
[49]Smith JR, Chang S F. Tools and techniques for color image retrieval. In: Proc of SPIE: Storage and Retrieval for Image and Video Database. San Jose, CA, 1996, 26(70)426-437
[50]Tversky A. Feature of similarity. Psychological Review, 1977, 84(4): 327-352
[51]倪明田 计算机图形学 北京：北京大学出版社 1999．11
[52]Essam A. El-kwae A Robust Framework for Content-Based Retrieval by Spatial Similarity in Image Database. Acm Transactions on Information Systems 1999.4 17(2) 174-198
[53]易江涛 硕士学位论文 基于内容的图象检索系统的设计与实现 中国科学技术大学 1999．4
[54]张海藩著 软件工程导论 北京：清华大学出版社 1999．5
[55]黄上腾 数据库原理 上海：上海交通大学出版社 1996．8
[56]Clifford A. Shaffer著 张铭等译 数据结构与算法分析(java版) 北京：电子工业出版社 2001．2
[57]Bruce Eckel著 京京工作室译 Java编程思想 北京：机械工业出版社 1999．4