链编码技术及区域标定自动机在图像处理中的应用
|
摘要
随着数字图像处理和模式识别技术的不断发展与成熟，图像的编码技术也得到了飞速发展，并以其优良的特性在图像处理领域得到越来越多的运用。
本文分析了数字图像处理技术的原理与发展，详细阐述了二值图像的链编码表示方法的原理及其特性。而后利用顶点链编码技术，在顾国庆和许彦冰发明的区域标定自动机的基础上，研究出了一种检测二值图像中区域角点的方法以及一套图像拼接的新算法。
其中的检测角点算法利用区域边界顶点链编码的特点可以准确检测到二值图像中边界线是直线的区域的角点。而图像拼接算法，针对灰度文字图像、无倾斜二值文字图像和倾斜的二值文字图像的拼接原理与实现进行了详述，三种图像拼接过程基本上都是利用链编码技术和区域标定自动机配合数学或图像处理方法获取到相邻重叠图像的匹配特征点进而实现拼接的。在进行二值图像拼接的研究过程中又提出了一种运用角点特征进行二值图像旋转的方法。
结合以上算法的技术和原理，利用Visual C++编程开发了算法实验系统，系统可实现二值图像的区域角点检测、灰度文字图像的拼接、二值文字图像的拼接以及只包含边界为直线段的区域的二值图像的旋转。
With the continual advances and maturation in digital image processing and pattern recognition, there is a speedy advance in image coding technique which is utilized more and more in the field of digital image processing by its fine characteristic .
After an analysis of the theory and advances of digital image processing, the thesis introduces in detail how a two-value image is expressed to the chain code form and the characteristic of the method. On the basis of region-labeling-automata invented by Gu Guo-qing and Xu Yan-bing, applying the vertex chain code technique, an algorithm for detecting and locating corners and a set of new stitching algorithm for document images are proposed.
Applying the characteristic of vertex chain code of region borderline, the algorithm for detecting corners can accurately detect the corners of the regions which borderlines are all beelines in two-value images .The stitching algorithms for images expatiate the stitching theory and technique for gray or two-value document images .The two-value document images include tilt and without tilt images. The stitching algorithms can fix the matching points of adjoining overlapping images using the method of vertex chain code and region-labeling-automata cooperating with maths or image processing technique. In the research of stitching for two-value images a method of circumgyrating for two-value images is given, which applying the characteristic of corners of regions.
Combining the above algorithms technique and theory, adopting the programme method of Visual C++,we develop a experiment system for above algorithms which function include detecting corners of region of two-value images, stitching gray and two-value document images and circumgyrating two-value images in which there only are region which borderlines are all beelines.
本文分析了数字图像处理技术的原理与发展，详细阐述了二值图像的链编码表示方法的原理及其特性。而后利用顶点链编码技术，在顾国庆和许彦冰发明的区域标定自动机的基础上，研究出了一种检测二值图像中区域角点的方法以及一套图像拼接的新算法。
其中的检测角点算法利用区域边界顶点链编码的特点可以准确检测到二值图像中边界线是直线的区域的角点。而图像拼接算法，针对灰度文字图像、无倾斜二值文字图像和倾斜的二值文字图像的拼接原理与实现进行了详述，三种图像拼接过程基本上都是利用链编码技术和区域标定自动机配合数学或图像处理方法获取到相邻重叠图像的匹配特征点进而实现拼接的。在进行二值图像拼接的研究过程中又提出了一种运用角点特征进行二值图像旋转的方法。
结合以上算法的技术和原理，利用Visual C++编程开发了算法实验系统，系统可实现二值图像的区域角点检测、灰度文字图像的拼接、二值文字图像的拼接以及只包含边界为直线段的区域的二值图像的旋转。
With the continual advances and maturation in digital image processing and pattern recognition, there is a speedy advance in image coding technique which is utilized more and more in the field of digital image processing by its fine characteristic .
After an analysis of the theory and advances of digital image processing, the thesis introduces in detail how a two-value image is expressed to the chain code form and the characteristic of the method. On the basis of region-labeling-automata invented by Gu Guo-qing and Xu Yan-bing, applying the vertex chain code technique, an algorithm for detecting and locating corners and a set of new stitching algorithm for document images are proposed.
Applying the characteristic of vertex chain code of region borderline, the algorithm for detecting corners can accurately detect the corners of the regions which borderlines are all beelines in two-value images .The stitching algorithms for images expatiate the stitching theory and technique for gray or two-value document images .The two-value document images include tilt and without tilt images. The stitching algorithms can fix the matching points of adjoining overlapping images using the method of vertex chain code and region-labeling-automata cooperating with maths or image processing technique. In the research of stitching for two-value images a method of circumgyrating for two-value images is given, which applying the characteristic of corners of regions.
Combining the above algorithms technique and theory, adopting the programme method of Visual C++,we develop a experiment system for above algorithms which function include detecting corners of region of two-value images, stitching gray and two-value document images and circumgyrating two-value images in which there only are region which borderlines are all beelines.
引文
[1] J B A Maintz, P A van den Elsen, M A Viergever. Evaluation of ridge seeking operaters for multimodality medical image matching. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1993. 18(4): 353-364
[2] Han M H, Jang D. The use of maximum curvature points for the recognition of partially occluded objects. Pattern Recognition. 1990.23:21-33
[3] H Wang, M Brady. Real-time comer detection algorithm for motion estimation. Image and Vision Computing. 1995.13(9): 695-703
[4] Freeman H. Computer processing of line-drawing images. Computing Surveys, 1974, 6 (1):57-97.
[5] Freeman H. On the encoding of arbitrary geometric configuration. IRE Trans, 1961, EC-10 (2):260-268.
[6] Freeman H. Techniques for the digital computer analysis of chain-encoded arbitrary plane curves. Proc. Natl. Elect. Conf., 1961, 17 (Oct.):421-432.
[7] Freeman H. A technique for the classification and recognition of geometric patterns. Proc. 3rd. Intl. Congress on Cybernetics Namtur, 1961 (Belgium):348-368.
[8] Sidhu G S, Boute R T. Property encoding: applications in binary picture encoding and boundary, following. IEEE Trans. Comp., 1972, C-21 (11):1206-1216.
[9] Merrill R. D. Representation of contours and regions for efficient computer search. Comm. ACM, 1973, 16(2):69-82.
[10] Bribiesca E. A new chain code. Pattern Recognition. 1999, 32: 235-251.
[11] 顾国庆，许彦冰．数字图像区域标定的方法．上海理工大学学报，Vol．23，No．4，2001：295-299．
[12] 许彦冰，顾国庆．Method to generate vertex chain code and the calculation of geometric quantities. Journal of Shanghai University, Vol. 5 Suppl. (Sep. 2001) 144-146.
[13] 顾国庆，陈可. Region-Labeling-Automata for Images in Square, Triangular and Hexagonal Lattice, Advances in Computer and Information Technology, The Journal of Three Dimensional Images, 17 (2003)151-165.
[14] 顾国庆，陈可．Region-Labeling-Automata for Images in Square, Triangular and Hexagonal Lattice, Advances in Computer and Information Technology, The Proceedings of 2002 International Conference on Computer and Information Technology, (Aizu-Wakamatsu City, Japan,Sept. 11-14, 2002) 134-138.
[15] Kenneth R. Castleman. Digital Image Processing. Prentice Hall, 1996. 北京：清华大学出版社，1998. 3:492-451.
[16] 沈庭芝，方子文．数字图像处理及模式识别．北京：北京理工大学出版社，1998：1-214．
[17] 王润生．图像理解．北京：国防科技大学出版社，1995：50-234
[18] 赵荣椿等．数字图像处理导论．西安：西北工业大学出版社，1999：30-231
[19] 林沛源，蒲和平．计算机图形图像处理应用技术．电子科技大学出版社，1997：1-63
[20] Nils J. Nilisson. Artificial Intelligence: A new Systhesis. 北京：机械工业出版社，2000
[21] 王碧泉，陈祖荫，模式识别理论、方法和应用．北京：地震出版社，1989：21-126
[22] 袁曾任．人工神经元网络及其应用．北京：清华大学出版社，广西：广西科学技术出版社，1999：1-27
[23] 沈清，汤霖．模式识别导论．北京：国防科技大学出版社，1990
[24] 傅京孙．模式识别及其应用．上海科技出版社，1983
[25] 张希平．我国汉字识别研究的进展．中文信息学报，1998：1-3
[26] Yihong Xu, George Nagy. Prototype Extraction and Adaptive OCR. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL.21, NO. 12, DECEMBER 1999
[27] Ari Gross,Longin Jan Latecki. Digital geometric methods in document image analysis.Pattern Recognition, 32(1999)407-424
[28] 罗希平，田捷，诸葛婴等．图像分割方法综述．模式识别和人工智能，1999：300-304
[29] 李国强，张薇，顾国庆．顶点链编码和区域面积的计算方法．上海理工大学学报，Vol．25，No．3．2003．
[30] Jacobs D.W,"Robust and efficient detection of salient convex groups",IEEE Trans on PAMI, Vol 18(1),pp 23-37,1996
[31] Kellman P. And Shipley T.,"A theory of visual interpolation in object perception",Cognitive Psychology, Vol 23,pp 141-221,1991
[32] Hsin.T and Hu W.C,"A rotationally invariant two phase scheme for comer detection",Pattern
Recognition,Vol 28,pp 819-828,1996
[33] Luo.B,Cross.A.D.J and Hacock..E.R,"Corner detection via topographic analysis of vector-potential",Pattern Recognition Letters,Vol 20,pp 635-650,1999
[34] Quddus.A and Fahmy.M.M, "Fast wavelet-based corner detection technique", Eletronics Letters,Vol 35(4),pp287-288,1999
[35] H L Beus ,S S H Tiu. An improved corner detection algorithm based on chain-coded plane curves[J]. PatternRecognition. 1983,20,291-276.
[36] L Kitchen ,A Rosedfeld. Gray-level corner detection[J]. Pattern Recognition Letters. 1982,1,95-102.
[37] 汪成为 灵境．(虚拟现实)技术的理论、实现及应用[M] 广西：广西科学技术出版社，1996
[38] 钟力，胡晓峰．重叠图像拼接算法[J] 中国图象图形学报，1998，3(5)：367～370
[39] 李志刚，纪玉波，薛金．边界重叠图像的一种快速拼接算法[J]计算机工程，2000，26(5)：37～38
[40] 肖忠，刘钊，史建雄．电子地图关键技术研究[J] 电子科技大学学报，1996，25(6)：10～11，42～43
[41] 沈建国，戴永刚：基于局部熵差的栅格地图拼接方法[J] 华东师范大学学报(自然科学版)，2002，4：48～53
[42] 吴伟．数字图像处理算法研究及其在名片识别系统中的应用，上海理工大学硕士学位论文，2003
[43] Peter Notton,Rob McGregor.开发Windows 95/Windows NT4应用程序．北京：清华大学出版社，1998
[44] Microsoft Corp. Microsoft Development Network Online Document
[45] 史文革．微机图像格式大全．北京：海洋大学出版社，1996
[46] David J．Kruglinski．Visual C++技术内幕(第四版)．北京：清华大学出版社，1999
[47] Davis Chapman,学用Visual C++6.0，北京：清华大学出版社，1999
[48] Microsoft Corp. Vistual C++ Online Help
[49] 何斌，马天予，王运坚，朱红莲．Visual C++数字图像处理．北京：人民邮电出版社，2001：4-674．
[50] 鸿志创作组．Visual C++ 5.0编程指南．北京：科学出版社，1998．4．
[51] 周长发．精通Visual C++图像编程．北京：电子工业出版
[2] Han M H, Jang D. The use of maximum curvature points for the recognition of partially occluded objects. Pattern Recognition. 1990.23:21-33
[3] H Wang, M Brady. Real-time comer detection algorithm for motion estimation. Image and Vision Computing. 1995.13(9): 695-703
[4] Freeman H. Computer processing of line-drawing images. Computing Surveys, 1974, 6 (1):57-97.
[5] Freeman H. On the encoding of arbitrary geometric configuration. IRE Trans, 1961, EC-10 (2):260-268.
[6] Freeman H. Techniques for the digital computer analysis of chain-encoded arbitrary plane curves. Proc. Natl. Elect. Conf., 1961, 17 (Oct.):421-432.
[7] Freeman H. A technique for the classification and recognition of geometric patterns. Proc. 3rd. Intl. Congress on Cybernetics Namtur, 1961 (Belgium):348-368.
[8] Sidhu G S, Boute R T. Property encoding: applications in binary picture encoding and boundary, following. IEEE Trans. Comp., 1972, C-21 (11):1206-1216.
[9] Merrill R. D. Representation of contours and regions for efficient computer search. Comm. ACM, 1973, 16(2):69-82.
[10] Bribiesca E. A new chain code. Pattern Recognition. 1999, 32: 235-251.
[11] 顾国庆，许彦冰．数字图像区域标定的方法．上海理工大学学报，Vol．23，No．4，2001：295-299．
[12] 许彦冰，顾国庆．Method to generate vertex chain code and the calculation of geometric quantities. Journal of Shanghai University, Vol. 5 Suppl. (Sep. 2001) 144-146.
[13] 顾国庆，陈可. Region-Labeling-Automata for Images in Square, Triangular and Hexagonal Lattice, Advances in Computer and Information Technology, The Journal of Three Dimensional Images, 17 (2003)151-165.
[14] 顾国庆，陈可．Region-Labeling-Automata for Images in Square, Triangular and Hexagonal Lattice, Advances in Computer and Information Technology, The Proceedings of 2002 International Conference on Computer and Information Technology, (Aizu-Wakamatsu City, Japan,Sept. 11-14, 2002) 134-138.
[15] Kenneth R. Castleman. Digital Image Processing. Prentice Hall, 1996. 北京：清华大学出版社，1998. 3:492-451.
[16] 沈庭芝，方子文．数字图像处理及模式识别．北京：北京理工大学出版社，1998：1-214．
[17] 王润生．图像理解．北京：国防科技大学出版社，1995：50-234
[18] 赵荣椿等．数字图像处理导论．西安：西北工业大学出版社，1999：30-231
[19] 林沛源，蒲和平．计算机图形图像处理应用技术．电子科技大学出版社，1997：1-63
[20] Nils J. Nilisson. Artificial Intelligence: A new Systhesis. 北京：机械工业出版社，2000
[21] 王碧泉，陈祖荫，模式识别理论、方法和应用．北京：地震出版社，1989：21-126
[22] 袁曾任．人工神经元网络及其应用．北京：清华大学出版社，广西：广西科学技术出版社，1999：1-27
[23] 沈清，汤霖．模式识别导论．北京：国防科技大学出版社，1990
[24] 傅京孙．模式识别及其应用．上海科技出版社，1983
[25] 张希平．我国汉字识别研究的进展．中文信息学报，1998：1-3
[26] Yihong Xu, George Nagy. Prototype Extraction and Adaptive OCR. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL.21, NO. 12, DECEMBER 1999
[27] Ari Gross,Longin Jan Latecki. Digital geometric methods in document image analysis.Pattern Recognition, 32(1999)407-424
[28] 罗希平，田捷，诸葛婴等．图像分割方法综述．模式识别和人工智能，1999：300-304
[29] 李国强，张薇，顾国庆．顶点链编码和区域面积的计算方法．上海理工大学学报，Vol．25，No．3．2003．
[30] Jacobs D.W,"Robust and efficient detection of salient convex groups",IEEE Trans on PAMI, Vol 18(1),pp 23-37,1996
[31] Kellman P. And Shipley T.,"A theory of visual interpolation in object perception",Cognitive Psychology, Vol 23,pp 141-221,1991
[32] Hsin.T and Hu W.C,"A rotationally invariant two phase scheme for comer detection",Pattern
Recognition,Vol 28,pp 819-828,1996
[33] Luo.B,Cross.A.D.J and Hacock..E.R,"Corner detection via topographic analysis of vector-potential",Pattern Recognition Letters,Vol 20,pp 635-650,1999
[34] Quddus.A and Fahmy.M.M, "Fast wavelet-based corner detection technique", Eletronics Letters,Vol 35(4),pp287-288,1999
[35] H L Beus ,S S H Tiu. An improved corner detection algorithm based on chain-coded plane curves[J]. PatternRecognition. 1983,20,291-276.
[36] L Kitchen ,A Rosedfeld. Gray-level corner detection[J]. Pattern Recognition Letters. 1982,1,95-102.
[37] 汪成为 灵境．(虚拟现实)技术的理论、实现及应用[M] 广西：广西科学技术出版社，1996
[38] 钟力，胡晓峰．重叠图像拼接算法[J] 中国图象图形学报，1998，3(5)：367～370
[39] 李志刚，纪玉波，薛金．边界重叠图像的一种快速拼接算法[J]计算机工程，2000，26(5)：37～38
[40] 肖忠，刘钊，史建雄．电子地图关键技术研究[J] 电子科技大学学报，1996，25(6)：10～11，42～43
[41] 沈建国，戴永刚：基于局部熵差的栅格地图拼接方法[J] 华东师范大学学报(自然科学版)，2002，4：48～53
[42] 吴伟．数字图像处理算法研究及其在名片识别系统中的应用，上海理工大学硕士学位论文，2003
[43] Peter Notton,Rob McGregor.开发Windows 95/Windows NT4应用程序．北京：清华大学出版社，1998
[44] Microsoft Corp. Microsoft Development Network Online Document
[45] 史文革．微机图像格式大全．北京：海洋大学出版社，1996
[46] David J．Kruglinski．Visual C++技术内幕(第四版)．北京：清华大学出版社，1999
[47] Davis Chapman,学用Visual C++6.0，北京：清华大学出版社，1999
[48] Microsoft Corp. Vistual C++ Online Help
[49] 何斌，马天予，王运坚，朱红莲．Visual C++数字图像处理．北京：人民邮电出版社，2001：4-674．
[50] 鸿志创作组．Visual C++ 5.0编程指南．北京：科学出版社，1998．4．
[51] 周长发．精通Visual C++图像编程．北京：电子工业出版
© 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 |