测井曲线矢量化方法及反演技术研究与实现
|
摘要
工程图纸矢量化是将纸介质工程图纸扫描输入计算机后,对所得的扫描图像加以分析、识别,最终重建其中的目标对象的过程。本文研究的测井解释成果图是进行油气勘探开发时使用的工程图纸。测井图纸的数字化将有利于数据的存储、管理、分析、共享和网络传输,因此测井曲线的矢量化就成为各油田的核心工作之一。目前通常采用数字化仪或国外矢量化软件来完成这一工作,但是速度较慢,效率较低。因此,如何把图纸上的测井曲线快速地自动识别和提取出来,就成为急待解决的问题。
本文针对这一要求开展了测井曲线自动识别与提取的研究工作,概括的讲,研究了以下内容:
1.设计了一种基于稀疏像素跟踪的曲线整体矢量化方法,该方法依据曲线的局部斜率跟踪曲线,在跟踪中动态调整跟踪方向和跟踪步长,并添加了搜索功能,从而既可以正确地穿越曲线上的相交和粘连部分,又可以跳过曲线的空隙。该方法包括五个基本过程,即确定起始点及跟踪方向、正常跟踪过程、矫正恢复过程、搜索过程、关键点提取过程。
2.提出了基于B样条插值的曲线反演策略,以B样条插值曲线表示原始曲线。B样条函数具有许多样条函数所没有的优点,如可以灵活地构成随意形状的曲线,移动某个节点只会影响临近的几段曲线等。基于B样条插值曲线反演策略首先根据曲线上的型值点反算B样条曲线的控制顶点,然后再进行B样条插值形成曲线。
基于以上关键技术并在以前工作的基础上进一步完善,最终完成了一个比较完整的测井图纸数字化系统,实现了图纸所有信息的数字化,包括图头、图例、曲线说明、图纸网格、曲线、解释成果表等,并把数字化信息存储到统一的数据库中。该系统还能够把存储到数据库中的数字化图纸信息提取出来,根据一定的反演算法实现图纸反演,即在新的图纸上重现原始图纸信息。
Engineering drawing vectorization is the process of analyzing and recognizing the image of engineering drawing, and finally rebuilding the graphic objects. Well log graph, which is used for the exploitation of oil and natural gas, is a kind of engineering drawing. Digitizing well log graph is favorable for storing, managing, analyzing, sharing and network transmitting well log information. Therefore vectorizing log curve becomes one of the key tasks for each oil field. Nowadays, vectorization of log curve is usually accomplished by the digitizer or foreign vectorizing software, but the vectorizing speed is relatively slow and the vectorizing efficiency is relatively low. How to recognize and pick up log curve from the well log graph automatically and rapidly, therefore, remains to be an emergent unresolved problem.
In order to solve the problem, this paper researches on automatic recognition and extraction of log curve that covers the following content:
1. We design a global curve vectorization method based on sparse pixel tracing. This method traces the curve by its local slope coefficient and automatically adjusts its tracing direction and step length. The added searching function of this method can not only traverse the intersecting and adhesive part of the curve correctly, but also leap over the curve gaps. This method involves five basic processes, namely start point and initial tracing direction confirmation process, the normal tracing process, correct and recovery process, searching process, and key point picking process.
2. We put forward the curve rebuilding strategy based on B-Spline interpolation and express the primary curve with B-spline curve. The B-spline function has many advantages that other spline functions don’t have, for example, forming curves of random shape, influencing only very few neighbor curved sections if moving a point. The strategy would first calculate the B-spline curve’s control points according to the given points, and then interpolate the B-spline to form the curve.
Based on the key techniques mentioned above and the previous study, this paper finally accomplishes a complete well log graph digital system, which can digitize all well log graph information, including graph head, legends, curve instructions, graph grids, curves, log interpretation table and so on, and all the digital information is stored in the database. This system can get digital well log graph information that has been stored in the database and rebuild the well log graph according to the rebuilding method, that is, to rebuild all graphic objects on a new image.
本文针对这一要求开展了测井曲线自动识别与提取的研究工作,概括的讲,研究了以下内容:
1.设计了一种基于稀疏像素跟踪的曲线整体矢量化方法,该方法依据曲线的局部斜率跟踪曲线,在跟踪中动态调整跟踪方向和跟踪步长,并添加了搜索功能,从而既可以正确地穿越曲线上的相交和粘连部分,又可以跳过曲线的空隙。该方法包括五个基本过程,即确定起始点及跟踪方向、正常跟踪过程、矫正恢复过程、搜索过程、关键点提取过程。
2.提出了基于B样条插值的曲线反演策略,以B样条插值曲线表示原始曲线。B样条函数具有许多样条函数所没有的优点,如可以灵活地构成随意形状的曲线,移动某个节点只会影响临近的几段曲线等。基于B样条插值曲线反演策略首先根据曲线上的型值点反算B样条曲线的控制顶点,然后再进行B样条插值形成曲线。
基于以上关键技术并在以前工作的基础上进一步完善,最终完成了一个比较完整的测井图纸数字化系统,实现了图纸所有信息的数字化,包括图头、图例、曲线说明、图纸网格、曲线、解释成果表等,并把数字化信息存储到统一的数据库中。该系统还能够把存储到数据库中的数字化图纸信息提取出来,根据一定的反演算法实现图纸反演,即在新的图纸上重现原始图纸信息。
Engineering drawing vectorization is the process of analyzing and recognizing the image of engineering drawing, and finally rebuilding the graphic objects. Well log graph, which is used for the exploitation of oil and natural gas, is a kind of engineering drawing. Digitizing well log graph is favorable for storing, managing, analyzing, sharing and network transmitting well log information. Therefore vectorizing log curve becomes one of the key tasks for each oil field. Nowadays, vectorization of log curve is usually accomplished by the digitizer or foreign vectorizing software, but the vectorizing speed is relatively slow and the vectorizing efficiency is relatively low. How to recognize and pick up log curve from the well log graph automatically and rapidly, therefore, remains to be an emergent unresolved problem.
In order to solve the problem, this paper researches on automatic recognition and extraction of log curve that covers the following content:
1. We design a global curve vectorization method based on sparse pixel tracing. This method traces the curve by its local slope coefficient and automatically adjusts its tracing direction and step length. The added searching function of this method can not only traverse the intersecting and adhesive part of the curve correctly, but also leap over the curve gaps. This method involves five basic processes, namely start point and initial tracing direction confirmation process, the normal tracing process, correct and recovery process, searching process, and key point picking process.
2. We put forward the curve rebuilding strategy based on B-Spline interpolation and express the primary curve with B-spline curve. The B-spline function has many advantages that other spline functions don’t have, for example, forming curves of random shape, influencing only very few neighbor curved sections if moving a point. The strategy would first calculate the B-spline curve’s control points according to the given points, and then interpolate the B-spline to form the curve.
Based on the key techniques mentioned above and the previous study, this paper finally accomplishes a complete well log graph digital system, which can digitize all well log graph information, including graph head, legends, curve instructions, graph grids, curves, log interpretation table and so on, and all the digital information is stored in the database. This system can get digital well log graph information that has been stored in the database and rebuild the well log graph according to the rebuilding method, that is, to rebuild all graphic objects on a new image.
引文
1孙文德.数字井筒开发研究调研论证报告.大庆油田第三采油厂信息中心技术报告. 2004: 1-3
2吴学毅.扫描图矢量化方法研究与实现.西安理工大学学报. 2003, 19(2): 177-181
3苏春光,徐杰,王鸿谷.地图及工程图纸的智能矢量化方法.小型微型计算机系统. 1996, 17(5): 6-11
4王加阳.王灿东.自动跟踪地图数字化.小型微型计算机系统. 2000, 8(11): 235-240
5 D. Dori and K. Toimbre. From Engineering Drawing to 3D CAD Models: Are We Ready Now? . Computer-Aided Design. 1995, 27(4): 243-254
6 S. H. Joseph and T. P. Pridmore. Knowledge-Directed Interpretation of Mechanical Engineering Drawings. IEEE Trans. Pattern Recognition and Machine Intelligence. 1992, 14(9): 928-940
7张李荪.工程图纸扫描输入及矢量化在水利工程设计中的应用.江西水利科技. 1997, 23(3): 155-157
8刘国翌.足球场地标志线的自动提取.计算机辅助设计与图形学学报. 2003, 15(7): 870-874
9杨杰,于德敏,许增朴,王永强.服装样片图像中直线与曲线矢量化的研究.天津工业大学学报. 2004, 23(1): 55-58
10曾志雄.浅谈数字化图形图像文件.现代计算机. 2000, 21(83): 53-55
11李新友,唐泽圣,张凤昌,黄汉文. TH-DAIMS 3. 0清华图纸自动输入与管理系统. 1997, 9(5): 451-455
12 C. S. Li, D. Cheng and C. Q. Zhang. A Platform to Integrate Well-Log Information Applications on Heterogeneous Environments. Proceedings of the 2nd International Conference on Information Technology for Applications, Harbin, 2004: 8-12
13孙文德,苗坤艳,杨立.动态监测数字化信息管理系统的开发研究调研论证报告.大庆油田第三采油厂信息中心技术报告. 2004: 5-10
14田玉敏,刘国景.光栅图形矢量化方法分析与评价.计算机应用研究. 2002, 3(1): 6-8
15丁伟东,袁景淇.工程图纸矢量化方法研究.计算机工程. 2006, 32(24): 157-159
16 Y. Zheng, H. Li and D. Doermann. A Parallel-Line Detection Algorithm Based on HMM Decoding. IEEE Trans. Pattern Analysis and Machine Intelligence. 2005, 27(5): 777-792
17 X. Hilaire and K. Tombre. Improving the Accuracy of Skeleton-Based Vectorization. Graphics Recognition--Algorithms and Applications, Kingston, 2002: 368-371
18 V. Nagasamy and N. A. Langrana. Engineering Drawing Processing and Vectorization System. Computer Vision, Graphics and Image Processing. 1990, 49(3): 379-397
19 O. Hori and S. Tanigawa. Raster-to-Vector Conversion by Line Fitting Based on Contours and Skeletons. Proceedings of the International Conference on Document Analysisi and Recognition, Tolyo, 1993: 272-281
20 L. Boatto. An Interpretation System for Land Register Maps. IEEE Computer. 1992, 25(7): 25-32
21 P. Vaxiviere and K. Tombre. Celesstin: CAD Conversion of Mechanical Drawings. IEEE Computer. 1992, 25(7): 46-54
22 G. Monagan and M. Roosli. Appropriate Base Representation Using a Run Graph. Proc. of ICDAR93, Tsukuba, 1993: 623-626
23王福顺.测井解释成果图矢量化方法的研究.哈尔滨工业大学硕士论文学位论文, 2005: 21-32
24 D. Dori, Y. Liang and I. Chai. Spare Pixel Recognition of Primitives in Engineering Drawings. Machine Vision and Applications. 1993, 6: 69-82
25 D. Dori. Orthogonal Zig-Zag: an Algorithm for Vectorizing Engineering Drawings Compared with Hough Transform. Advances in Engineering Software. 1997, 28(1): 11-24
26 W. Liu and D. Dori. Sparse Pixel Tracking: A Fast Vectorization Algorithm Applied to Engineering Drawings. Proc. ICPR96 Volume III, Robotics and Applications, Vienna, 1996: 808-811
27 D. Dori. Sparse Pixel Vectorization: An Algorithm and Its Performance Evaluation. IEEE Trans. on Pattern Analysis and Machine Intelligence. 1999, 21(3): 202-215
28钟必能.基于逐步简化的测井解释成果图矢量化的研究与实现.哈尔滨工业大学硕士学位论文. 2006: 7-8
29 Donald Hearn and M. Pauline Baker. Computer Graphics with OpenGL.蔡士杰,宋继强,蔡敏译.第三版.电子工业出版社, 2005: 364-371
30雷开彬,张为群,马志霞.一种新的均匀样条曲线曲面设计方法.计算机科学. 2006, 33(9): 225-226
31方忆湘,刘文学.基于几何特的三次均匀B样条曲线构造描述.工程图学学报. 2006, 2(2): 96-99
32林意,熊汉伟,骆少明,张湘伟.过控制顶点的B样条曲线.江南大学学报. 2003, 2(6): 553-555
33 P. K. Sahoo, S. Soltani and A. K. C. Wong. A Survey of Thresholding Techniques. CVGIP. 1988, 41: 233-260
34 J. R. Parker, C. Jennings and A. G. Salkauskas. Thresholding Using an Illumination Model. IEEE Trans on PAMI. 1993, 15(7): 270-273
35 Rafael, C. Gonzalez and E. Richard. Digital Image Processing. Publishing House of Electronics Industry, 2002: 489-495
36罗峰,余艳梅,吴晓红,滕奇志.曲线矢量化中的背景去噪方法.计算机应用. 2005, 25(4): 929-932
37刘波,余艳梅,罗峰,滕奇志.数字图像处理在测井曲线矢量化中的应用. 2005, 31(1): 148-151
38 R.M. Haralik. Performance Characterization in Image Analysis-Thinning, A Case in Point. Pattern Recognition Letters. 1992, 13(1): 5-12
39 B. Kong. A Benchmark: Performance Evaluation of Dashed-Line Detection Algorithms. Graphics Recognition--Methods and Application, Springer, 1996: 270-285
40 O. Hori and D. Doermann. Quantitative Measurement of the Performance of Raster-to-Vector Conversion Algorithms. Graphics Recognition--Methods and Application, Springer, 1996: 57-68
41 W. Liu and D. Dori. A Protocol for Performance Evaluation of Line Detection Algorithms. Machine Vision Applications. 1997, 9(5): 240-250
42 I. T. Phillips, J. Liang, A. Chhabra and R. M. Haralick. A Performance Evaluation Protocol for Graphics Recognition Systems. Graphics Recognition--Algorithms and Systems, Springer, 1998: 372-389
2吴学毅.扫描图矢量化方法研究与实现.西安理工大学学报. 2003, 19(2): 177-181
3苏春光,徐杰,王鸿谷.地图及工程图纸的智能矢量化方法.小型微型计算机系统. 1996, 17(5): 6-11
4王加阳.王灿东.自动跟踪地图数字化.小型微型计算机系统. 2000, 8(11): 235-240
5 D. Dori and K. Toimbre. From Engineering Drawing to 3D CAD Models: Are We Ready Now? . Computer-Aided Design. 1995, 27(4): 243-254
6 S. H. Joseph and T. P. Pridmore. Knowledge-Directed Interpretation of Mechanical Engineering Drawings. IEEE Trans. Pattern Recognition and Machine Intelligence. 1992, 14(9): 928-940
7张李荪.工程图纸扫描输入及矢量化在水利工程设计中的应用.江西水利科技. 1997, 23(3): 155-157
8刘国翌.足球场地标志线的自动提取.计算机辅助设计与图形学学报. 2003, 15(7): 870-874
9杨杰,于德敏,许增朴,王永强.服装样片图像中直线与曲线矢量化的研究.天津工业大学学报. 2004, 23(1): 55-58
10曾志雄.浅谈数字化图形图像文件.现代计算机. 2000, 21(83): 53-55
11李新友,唐泽圣,张凤昌,黄汉文. TH-DAIMS 3. 0清华图纸自动输入与管理系统. 1997, 9(5): 451-455
12 C. S. Li, D. Cheng and C. Q. Zhang. A Platform to Integrate Well-Log Information Applications on Heterogeneous Environments. Proceedings of the 2nd International Conference on Information Technology for Applications, Harbin, 2004: 8-12
13孙文德,苗坤艳,杨立.动态监测数字化信息管理系统的开发研究调研论证报告.大庆油田第三采油厂信息中心技术报告. 2004: 5-10
14田玉敏,刘国景.光栅图形矢量化方法分析与评价.计算机应用研究. 2002, 3(1): 6-8
15丁伟东,袁景淇.工程图纸矢量化方法研究.计算机工程. 2006, 32(24): 157-159
16 Y. Zheng, H. Li and D. Doermann. A Parallel-Line Detection Algorithm Based on HMM Decoding. IEEE Trans. Pattern Analysis and Machine Intelligence. 2005, 27(5): 777-792
17 X. Hilaire and K. Tombre. Improving the Accuracy of Skeleton-Based Vectorization. Graphics Recognition--Algorithms and Applications, Kingston, 2002: 368-371
18 V. Nagasamy and N. A. Langrana. Engineering Drawing Processing and Vectorization System. Computer Vision, Graphics and Image Processing. 1990, 49(3): 379-397
19 O. Hori and S. Tanigawa. Raster-to-Vector Conversion by Line Fitting Based on Contours and Skeletons. Proceedings of the International Conference on Document Analysisi and Recognition, Tolyo, 1993: 272-281
20 L. Boatto. An Interpretation System for Land Register Maps. IEEE Computer. 1992, 25(7): 25-32
21 P. Vaxiviere and K. Tombre. Celesstin: CAD Conversion of Mechanical Drawings. IEEE Computer. 1992, 25(7): 46-54
22 G. Monagan and M. Roosli. Appropriate Base Representation Using a Run Graph. Proc. of ICDAR93, Tsukuba, 1993: 623-626
23王福顺.测井解释成果图矢量化方法的研究.哈尔滨工业大学硕士论文学位论文, 2005: 21-32
24 D. Dori, Y. Liang and I. Chai. Spare Pixel Recognition of Primitives in Engineering Drawings. Machine Vision and Applications. 1993, 6: 69-82
25 D. Dori. Orthogonal Zig-Zag: an Algorithm for Vectorizing Engineering Drawings Compared with Hough Transform. Advances in Engineering Software. 1997, 28(1): 11-24
26 W. Liu and D. Dori. Sparse Pixel Tracking: A Fast Vectorization Algorithm Applied to Engineering Drawings. Proc. ICPR96 Volume III, Robotics and Applications, Vienna, 1996: 808-811
27 D. Dori. Sparse Pixel Vectorization: An Algorithm and Its Performance Evaluation. IEEE Trans. on Pattern Analysis and Machine Intelligence. 1999, 21(3): 202-215
28钟必能.基于逐步简化的测井解释成果图矢量化的研究与实现.哈尔滨工业大学硕士学位论文. 2006: 7-8
29 Donald Hearn and M. Pauline Baker. Computer Graphics with OpenGL.蔡士杰,宋继强,蔡敏译.第三版.电子工业出版社, 2005: 364-371
30雷开彬,张为群,马志霞.一种新的均匀样条曲线曲面设计方法.计算机科学. 2006, 33(9): 225-226
31方忆湘,刘文学.基于几何特的三次均匀B样条曲线构造描述.工程图学学报. 2006, 2(2): 96-99
32林意,熊汉伟,骆少明,张湘伟.过控制顶点的B样条曲线.江南大学学报. 2003, 2(6): 553-555
33 P. K. Sahoo, S. Soltani and A. K. C. Wong. A Survey of Thresholding Techniques. CVGIP. 1988, 41: 233-260
34 J. R. Parker, C. Jennings and A. G. Salkauskas. Thresholding Using an Illumination Model. IEEE Trans on PAMI. 1993, 15(7): 270-273
35 Rafael, C. Gonzalez and E. Richard. Digital Image Processing. Publishing House of Electronics Industry, 2002: 489-495
36罗峰,余艳梅,吴晓红,滕奇志.曲线矢量化中的背景去噪方法.计算机应用. 2005, 25(4): 929-932
37刘波,余艳梅,罗峰,滕奇志.数字图像处理在测井曲线矢量化中的应用. 2005, 31(1): 148-151
38 R.M. Haralik. Performance Characterization in Image Analysis-Thinning, A Case in Point. Pattern Recognition Letters. 1992, 13(1): 5-12
39 B. Kong. A Benchmark: Performance Evaluation of Dashed-Line Detection Algorithms. Graphics Recognition--Methods and Application, Springer, 1996: 270-285
40 O. Hori and D. Doermann. Quantitative Measurement of the Performance of Raster-to-Vector Conversion Algorithms. Graphics Recognition--Methods and Application, Springer, 1996: 57-68
41 W. Liu and D. Dori. A Protocol for Performance Evaluation of Line Detection Algorithms. Machine Vision Applications. 1997, 9(5): 240-250
42 I. T. Phillips, J. Liang, A. Chhabra and R. M. Haralick. A Performance Evaluation Protocol for Graphics Recognition Systems. Graphics Recognition--Algorithms and Systems, Springer, 1998: 372-389