数字图像辅助激光点云特征提取研究
|
摘要
随着地面、车载、机载激光扫描采集系统技术的成熟,国内外越来越多的研究人员开始研究基于激光点云的物体三维建模。点云特征提取作为基于激光点云三维建模的一个重要环节,也逐渐成为其中的一个研究热点。
针对已有点云特征提取算法不易于理解、实现难度大、数学运算复杂等缺点,综合数字图像和激光点云各自的优点,本文提出数字图像辅助激光点云特征提取方法,并给出详细的方法流程及实验结果。核心思想是将点云中的点映射到对应的二维图像中的像素,然后从图像中提取目标物体的特征,将特征所包含的像素根据对应关系找到在点云中的对应点,之后对这些点进行曲线和曲面拟合,由此得到点云的特征。需要解决的关键问题就是点云和对应二维图像的对应关系的建立,即点云和图像的配准。为了区别,本文将点云对应的由CCD相机获取的图像称为CCD图像。利用强度图像能够反映原始点云大部分特征的特点,将点云与CCD图像同名点的提取转换为点云强度图像与CCD图像的同名点提取。针对CCD图像和点云强度图像灰度差异大而不能使用传统的基于灰度信息的配准问题,引入医学图像配准中常用的基于互信息的图像配准算法,针对特征点的提取使用改进的自适应阈值Harris角点提取算法。利用共线条件方程解决点云和CCD图像的配准问题,通过曲线拟合和曲面拟合实现点云特征的精确提取。实验结果表明本文方法能够比较准确地提取点云特征,并且实现难度小、没有大量复杂的数学运算,具有一定的应用价值。
As the terrestrial, vehicles borne, air borne laser scanning data acquisition system technologies matured, more and more domestic and foreign researchers began to study on 3D modeling based on point cloud. Point cloud feature extraction as an important link in 3D laser scanning modeling, has gradually become one of the research hotspots.
It is difficult to understand and realize the existed algorithms of point cloud feature extraction. To deal with these disadvantages, it puts forward a new algorithm to extract features from point cloud indirectly in this paper which named digital image assists point cloud feature extraction, and a detailed process of this algorithm and results of the experiment have been given. The core idea of this algorithm is mapping the points in the point cloud to the pixels in the 2D image,then get features of target object from 2D image, According to the registration relationship between the pixels in 2D image and points in point cloud, the points in point cloud compose the features can be found, using curve and surface fitting to make features more accurate, these curves or surfaces are the features extracted from point cloud. The key problem that needs to be resolved is the establishment of registration relationship between 2D image and point cloud. To avoid confusion, the 2D image acquired by CCD camera is called CCD image here. Since the intensity image of the point cloud can reflects most features of the original point cloud, the registration between point cloud and CCD image can be divided into two steps. First, register the CCD image and intensity image, in view of grayscale variation between CCD image and intensity image is too great to register them by use of traditional image register algorithms which based on gray information, the image register algorithm based on mutual information usually applied in medical image registration is introduced. Improved adaptive threshold Harris corner detection algorithm is used to extract feature points from images. Second, register the CCD image and point cloud, the collinear condition equation is introduced to solve this problem. Since registration relationship between CCD image and point cloud has been established, features extracted from CCD image can be mapped into the point cloud, with curve and surface fitting, the curves or surfaces fitted in point cloud are the features of point cloud. The experiment results show that this algorithm can extract features from point cloud more accurately and with less operations, can be applied in some solutions.
针对已有点云特征提取算法不易于理解、实现难度大、数学运算复杂等缺点,综合数字图像和激光点云各自的优点,本文提出数字图像辅助激光点云特征提取方法,并给出详细的方法流程及实验结果。核心思想是将点云中的点映射到对应的二维图像中的像素,然后从图像中提取目标物体的特征,将特征所包含的像素根据对应关系找到在点云中的对应点,之后对这些点进行曲线和曲面拟合,由此得到点云的特征。需要解决的关键问题就是点云和对应二维图像的对应关系的建立,即点云和图像的配准。为了区别,本文将点云对应的由CCD相机获取的图像称为CCD图像。利用强度图像能够反映原始点云大部分特征的特点,将点云与CCD图像同名点的提取转换为点云强度图像与CCD图像的同名点提取。针对CCD图像和点云强度图像灰度差异大而不能使用传统的基于灰度信息的配准问题,引入医学图像配准中常用的基于互信息的图像配准算法,针对特征点的提取使用改进的自适应阈值Harris角点提取算法。利用共线条件方程解决点云和CCD图像的配准问题,通过曲线拟合和曲面拟合实现点云特征的精确提取。实验结果表明本文方法能够比较准确地提取点云特征,并且实现难度小、没有大量复杂的数学运算,具有一定的应用价值。
As the terrestrial, vehicles borne, air borne laser scanning data acquisition system technologies matured, more and more domestic and foreign researchers began to study on 3D modeling based on point cloud. Point cloud feature extraction as an important link in 3D laser scanning modeling, has gradually become one of the research hotspots.
It is difficult to understand and realize the existed algorithms of point cloud feature extraction. To deal with these disadvantages, it puts forward a new algorithm to extract features from point cloud indirectly in this paper which named digital image assists point cloud feature extraction, and a detailed process of this algorithm and results of the experiment have been given. The core idea of this algorithm is mapping the points in the point cloud to the pixels in the 2D image,then get features of target object from 2D image, According to the registration relationship between the pixels in 2D image and points in point cloud, the points in point cloud compose the features can be found, using curve and surface fitting to make features more accurate, these curves or surfaces are the features extracted from point cloud. The key problem that needs to be resolved is the establishment of registration relationship between 2D image and point cloud. To avoid confusion, the 2D image acquired by CCD camera is called CCD image here. Since the intensity image of the point cloud can reflects most features of the original point cloud, the registration between point cloud and CCD image can be divided into two steps. First, register the CCD image and intensity image, in view of grayscale variation between CCD image and intensity image is too great to register them by use of traditional image register algorithms which based on gray information, the image register algorithm based on mutual information usually applied in medical image registration is introduced. Improved adaptive threshold Harris corner detection algorithm is used to extract feature points from images. Second, register the CCD image and point cloud, the collinear condition equation is introduced to solve this problem. Since registration relationship between CCD image and point cloud has been established, features extracted from CCD image can be mapped into the point cloud, with curve and surface fitting, the curves or surfaces fitted in point cloud are the features of point cloud. The experiment results show that this algorithm can extract features from point cloud more accurately and with less operations, can be applied in some solutions.
引文
[1]陈曦.反求工程中基于点云的特征挖掘技术研究[D]：[博士学位论文].杭州：浙江大学,2005
[2]杨振羽.基于点表示几何体的造型技术[D]：[博士学位论文].杭州：浙江大学,2004
[3]Eric K Forkuo, Bruce King. Automatic Fusion of Photogrammetric Imagery and Laser Scanner Point Clouds [D]. Department of Land Surveying & Geo-Informatics The Hong Kong Polytechnic University Hung Hom, Hong Kong,2004
[4]Alharthy, A. and J. Bethel,2002. Building extraction and reconstruction from LIDAR data[C]. Proceedings of ASPRS annual conference,18-26, Washington
[5]路兴昌.基于激光扫描数据的三维场景仿真[J].系统仿真学报,2006,18(1)：176-178
[6]李清泉,杨必胜.三维空间数据的实时获取、建模与可视化[M].武汉：武汉大学出版社,2003
[7]邓非,张祖勋,张剑清.利用激光扫描和数码相机进行古建筑三维重建研究[J].测绘科学,2007,32(2)：29-30
[8]余明,丁辰,过静.激光三维扫描技术用于古建筑测绘的研究[J].测绘科学,2004,(5)：69-70
[9]Liu Jie, Zhang Jianqing et al. Cultural Relic 3D Reconstruction from Digital Images and Laser Point Clouds [C].2008 Congress on Image and Signal Processing,2008
[10]Blais, F. Review of 20 Years of Range Sensor Development [J]. Journal of Electronic Imaging,2004
[11]Jianbing Huang, Chia-Hsiang Menq. Automatic Data Segmentation for Geometric Feature Extraction From Unorganized 3-D Coordinate Points [J].IEEE TRANSACTIONS ON ROBOTICS AND AUTOMATION,2001,17 (3):268 279
[12]SABRY F. El-Hakim, CLAUS Brenner et al. A multi-sensor approach to creating accurate virtual environment [J]. ISPRS Journal of Photogrammetric & Remote Sensing,1998, (53):379-391
[13]李长春,薛华柱等.三维激光扫描在建筑物模型构建中的研究与实现[J].河南理工大学学报(自然科学版),2008,27(2)：193-199
[14]栾悉道,应龙等.三维建模技术研究进展[J].计算机科学,2008,135(12)：208-210
[15]田鹏.地理场景中点云特征提取与简化研究[D]：[硕士学位论文].南京：南京师范大学,2008
[16]Milroy M J, Bradley C et al. Segmentation of a wrap-around model using an active contour[J]. Computer Aided Design 1997,29(4):299-320
[17]Gumhold S, Wang X et al. Feature extraction from point clouds[C].10th International Meshing Roundtable,2001:293-305
[18]Pauly M, Gross M et al. Efficient Simplification of Point-Sampled Surfaces[J]. In Proceedings Visualization 2002,:163-170
[19]柯映林,陈曦.基于4DShepard曲面的点云曲率估算[J].浙江大学学报(工学版),2005,39(6)：761-764
[20]Dinesh M, Ryosuke S. Auto-extraction of Urban Features from Vehicle Borne Laser data[C]. Symposium on Geospatial Theory, Processing and Applications, Ottawa. 2002
[21]肖华.网格重构及特征提取技术研究[D]：[硕士学位论文].杭州：浙江大学,2010
[22]王瑶.从三维点云数据中提取物体特征点的研究[D]：[硕士学位论文].兰州：兰州大学,2010
[23]刘伟军,孙玉文等.逆向工程原理、方法及应用[M].北京：机械工业出版社,2009
[24]毛方儒,王磊.三维激光扫描测量技术[J].宇航计测技术,2005,25(2)：1-6
[25]蔡润彬.地面激光扫描数据后处理若干关键技术研究[D]：[博士学位论文].上海：同济大学,2008
[26]柯映林,范树迁.基于点云的边界特征直接提取技术[J].机械工程学报,2004,40(9)：116-120
[27]孙殿柱,孙肖霞等.反求工程中散乱数据边界特征自动提取算法[J].机械设计,2006,23：213-216
[28]孙殿柱,范志先等.散乱数据点云型面特征提取算法研究[J].机械设计与研究,2007,23(4)：77-81
[29]Ruwen Schnabel Roland Wahl et al. Efficient RANSAC for Point-Cloud Shape Detection[J].The Euro graphics Association and Blackwell Publishing,2007
[30]Haala N, Brenner C. Vitual City Models from Laser Altimeter and 2D Map Data[J]. PE&RS'99,1999,65(7):787-795.
[31]陈远.激光和CCD数据融合的三维重建关键技术研究[D]：[硕士学位论文]. 南昌：南昌航空大学,2008
[32]刘满仓.三维城市建模中正射影像图的处理技巧[J].三晋测绘,2003,10(2)：28-29
[33]胡鑫,习俊通,金烨.基于图像法的点云数据边界自动提取[J].上海交通大学学报,2002,36(8)：1118-1120
[34]邹万红,陈志杨,叶修梓等.一种新的点云数据特征骨架提取方法[J].浙江大学学报(工学版),2008,42(12)：2103-2107
[35]尤红建,张世强.组合CCD图像和稀疏激光测距数据的建筑物三维信息提取[J].光学精密工程,2006,14(2)：297-302
[36]陈远,陈震等. 基于CCD图像的点云区域分割方法[J]. 南昌航空工业学院学报(自然科学版),2007,21(1)：40-42
[37]湛金辉,陈震等.基于CCD图像的激光点云数据边界提取法[J].山西电子技术,2008,(3)：3-6
[38]张帆,黄先锋等.激光扫描与光学影像数据配准的研究进展[J].测绘通报,2008,(2)： 7-10
[39]冈萨雷斯著,阮秋琦译.数字图像处理(第二版[M]).北京：电子工业出版社,2006
[40]Brown LG. A survey of image registration techniques. ACM Computing surveys, 1992,24(4):325-376
[41]F Maes, D Vandermeulen et al. Medical image registration using mutual information [J].Proceedings of the IEEE.2003,91(10):699-1721
[42]吕煊.基于角点特征和最大互信息的多模医学图像配准[D]：[硕士学位论文].济南：山东师范大学,2008
[43]周永新,罗述谦.基于形状特征点最大互信息的医学图像配准[J].计算机辅助设计与图形学学报,2002,14(7)：654-658
[44]钟健.基于对极几何的图像匹配研究[D]：[硕士学位论文].长沙：中南大学,2010
[45]Nicholas Shorter, Talis Kasparis. Autonomous registration of LiDAR data to single aerial image [J]. Geoscience and Remote sensing Symposium, (IGARSS),2008, 5(7-11):216-219
[46]张俊峰,吴颖超等.基于特征点的数字图像配准[J].河南科学,25(6)：992-994
[47]李洋.基于Harris角点的Lidar图像与遥感图像的自动配准[D]：[硕士学位论文].合肥：合肥工业大学,2010
[48]刘睿,王锋等.基于小波变换多尺度Harris角点检测算法[J].微计算机信息.2009,6(3)：244-245
[49]于起峰,尚洋.摄影测量学原理与应用研究[M].北京：科学出版社,2009
[50]李牧,闫继红等.自适应Canny算子边缘检测技术[J].哈尔滨工程大学学报,2007,28(9)：1002-1007
[51]John Canny. A computational approach to edge detection [J].IEEE Transaction Pattern Analysis,1986,PAMI-8:679-698
[52]谢志孟,高向东.基于Canny算子的焊缝图像边缘提取技术[J].焊接学报,2006,27(1)：29-34
[53]彭芳瑜,周济等.基于最小二乘法的曲面生成算法研究[J].工程图学学报,1999,(3)：41-47
[54]Gu P, Yan X. Neural Network Approach to the Reconstruction of Freeform Surfaces for Reverse Engineering. Computer-aided Design,1995,27(1):59-64
[55]Frey P J. Generation and Adaptation of Computational Surface Meshes from Discrete Anatomical Data. International Journal for Numerical Methods in Engineering,2004,60(2):1049-1074
[56]孙家广等.计算机图形学(第三版).北京：清华大学出版社,1998
[57]许智钦,孙长库.3D逆向工程技术.北京：中国计量出版社,2002
[58]Ruud MBolle, Baba C Vemuri. On three-dimensional surface reconstruction. IEEE PAMI,1991,13(1):1-12
[59]仇芝.三维重建的匹配技术[D].南京理工大学,2004
[60]ILRIS-3D Operation Manual [M]. Optech Incorporated Industrial & 3D Imaging Divsion.2006
[61]赵艳平,高明等.基于PolyWorks的逆向工程数据处理[J].计算机应用技术,2005,32(10)：38-40
[62]许建中,马利庄.改进的Delaunay三角网渐次插入生成算法[J].计算机工程,2008,34(17)：254-256
[2]杨振羽.基于点表示几何体的造型技术[D]：[博士学位论文].杭州：浙江大学,2004
[3]Eric K Forkuo, Bruce King. Automatic Fusion of Photogrammetric Imagery and Laser Scanner Point Clouds [D]. Department of Land Surveying & Geo-Informatics The Hong Kong Polytechnic University Hung Hom, Hong Kong,2004
[4]Alharthy, A. and J. Bethel,2002. Building extraction and reconstruction from LIDAR data[C]. Proceedings of ASPRS annual conference,18-26, Washington
[5]路兴昌.基于激光扫描数据的三维场景仿真[J].系统仿真学报,2006,18(1)：176-178
[6]李清泉,杨必胜.三维空间数据的实时获取、建模与可视化[M].武汉：武汉大学出版社,2003
[7]邓非,张祖勋,张剑清.利用激光扫描和数码相机进行古建筑三维重建研究[J].测绘科学,2007,32(2)：29-30
[8]余明,丁辰,过静.激光三维扫描技术用于古建筑测绘的研究[J].测绘科学,2004,(5)：69-70
[9]Liu Jie, Zhang Jianqing et al. Cultural Relic 3D Reconstruction from Digital Images and Laser Point Clouds [C].2008 Congress on Image and Signal Processing,2008
[10]Blais, F. Review of 20 Years of Range Sensor Development [J]. Journal of Electronic Imaging,2004
[11]Jianbing Huang, Chia-Hsiang Menq. Automatic Data Segmentation for Geometric Feature Extraction From Unorganized 3-D Coordinate Points [J].IEEE TRANSACTIONS ON ROBOTICS AND AUTOMATION,2001,17 (3):268 279
[12]SABRY F. El-Hakim, CLAUS Brenner et al. A multi-sensor approach to creating accurate virtual environment [J]. ISPRS Journal of Photogrammetric & Remote Sensing,1998, (53):379-391
[13]李长春,薛华柱等.三维激光扫描在建筑物模型构建中的研究与实现[J].河南理工大学学报(自然科学版),2008,27(2)：193-199
[14]栾悉道,应龙等.三维建模技术研究进展[J].计算机科学,2008,135(12)：208-210
[15]田鹏.地理场景中点云特征提取与简化研究[D]：[硕士学位论文].南京：南京师范大学,2008
[16]Milroy M J, Bradley C et al. Segmentation of a wrap-around model using an active contour[J]. Computer Aided Design 1997,29(4):299-320
[17]Gumhold S, Wang X et al. Feature extraction from point clouds[C].10th International Meshing Roundtable,2001:293-305
[18]Pauly M, Gross M et al. Efficient Simplification of Point-Sampled Surfaces[J]. In Proceedings Visualization 2002,:163-170
[19]柯映林,陈曦.基于4DShepard曲面的点云曲率估算[J].浙江大学学报(工学版),2005,39(6)：761-764
[20]Dinesh M, Ryosuke S. Auto-extraction of Urban Features from Vehicle Borne Laser data[C]. Symposium on Geospatial Theory, Processing and Applications, Ottawa. 2002
[21]肖华.网格重构及特征提取技术研究[D]：[硕士学位论文].杭州：浙江大学,2010
[22]王瑶.从三维点云数据中提取物体特征点的研究[D]：[硕士学位论文].兰州：兰州大学,2010
[23]刘伟军,孙玉文等.逆向工程原理、方法及应用[M].北京：机械工业出版社,2009
[24]毛方儒,王磊.三维激光扫描测量技术[J].宇航计测技术,2005,25(2)：1-6
[25]蔡润彬.地面激光扫描数据后处理若干关键技术研究[D]：[博士学位论文].上海：同济大学,2008
[26]柯映林,范树迁.基于点云的边界特征直接提取技术[J].机械工程学报,2004,40(9)：116-120
[27]孙殿柱,孙肖霞等.反求工程中散乱数据边界特征自动提取算法[J].机械设计,2006,23：213-216
[28]孙殿柱,范志先等.散乱数据点云型面特征提取算法研究[J].机械设计与研究,2007,23(4)：77-81
[29]Ruwen Schnabel Roland Wahl et al. Efficient RANSAC for Point-Cloud Shape Detection[J].The Euro graphics Association and Blackwell Publishing,2007
[30]Haala N, Brenner C. Vitual City Models from Laser Altimeter and 2D Map Data[J]. PE&RS'99,1999,65(7):787-795.
[31]陈远.激光和CCD数据融合的三维重建关键技术研究[D]：[硕士学位论文]. 南昌：南昌航空大学,2008
[32]刘满仓.三维城市建模中正射影像图的处理技巧[J].三晋测绘,2003,10(2)：28-29
[33]胡鑫,习俊通,金烨.基于图像法的点云数据边界自动提取[J].上海交通大学学报,2002,36(8)：1118-1120
[34]邹万红,陈志杨,叶修梓等.一种新的点云数据特征骨架提取方法[J].浙江大学学报(工学版),2008,42(12)：2103-2107
[35]尤红建,张世强.组合CCD图像和稀疏激光测距数据的建筑物三维信息提取[J].光学精密工程,2006,14(2)：297-302
[36]陈远,陈震等. 基于CCD图像的点云区域分割方法[J]. 南昌航空工业学院学报(自然科学版),2007,21(1)：40-42
[37]湛金辉,陈震等.基于CCD图像的激光点云数据边界提取法[J].山西电子技术,2008,(3)：3-6
[38]张帆,黄先锋等.激光扫描与光学影像数据配准的研究进展[J].测绘通报,2008,(2)： 7-10
[39]冈萨雷斯著,阮秋琦译.数字图像处理(第二版[M]).北京：电子工业出版社,2006
[40]Brown LG. A survey of image registration techniques. ACM Computing surveys, 1992,24(4):325-376
[41]F Maes, D Vandermeulen et al. Medical image registration using mutual information [J].Proceedings of the IEEE.2003,91(10):699-1721
[42]吕煊.基于角点特征和最大互信息的多模医学图像配准[D]：[硕士学位论文].济南：山东师范大学,2008
[43]周永新,罗述谦.基于形状特征点最大互信息的医学图像配准[J].计算机辅助设计与图形学学报,2002,14(7)：654-658
[44]钟健.基于对极几何的图像匹配研究[D]：[硕士学位论文].长沙：中南大学,2010
[45]Nicholas Shorter, Talis Kasparis. Autonomous registration of LiDAR data to single aerial image [J]. Geoscience and Remote sensing Symposium, (IGARSS),2008, 5(7-11):216-219
[46]张俊峰,吴颖超等.基于特征点的数字图像配准[J].河南科学,25(6)：992-994
[47]李洋.基于Harris角点的Lidar图像与遥感图像的自动配准[D]：[硕士学位论文].合肥：合肥工业大学,2010
[48]刘睿,王锋等.基于小波变换多尺度Harris角点检测算法[J].微计算机信息.2009,6(3)：244-245
[49]于起峰,尚洋.摄影测量学原理与应用研究[M].北京：科学出版社,2009
[50]李牧,闫继红等.自适应Canny算子边缘检测技术[J].哈尔滨工程大学学报,2007,28(9)：1002-1007
[51]John Canny. A computational approach to edge detection [J].IEEE Transaction Pattern Analysis,1986,PAMI-8:679-698
[52]谢志孟,高向东.基于Canny算子的焊缝图像边缘提取技术[J].焊接学报,2006,27(1)：29-34
[53]彭芳瑜,周济等.基于最小二乘法的曲面生成算法研究[J].工程图学学报,1999,(3)：41-47
[54]Gu P, Yan X. Neural Network Approach to the Reconstruction of Freeform Surfaces for Reverse Engineering. Computer-aided Design,1995,27(1):59-64
[55]Frey P J. Generation and Adaptation of Computational Surface Meshes from Discrete Anatomical Data. International Journal for Numerical Methods in Engineering,2004,60(2):1049-1074
[56]孙家广等.计算机图形学(第三版).北京：清华大学出版社,1998
[57]许智钦,孙长库.3D逆向工程技术.北京：中国计量出版社,2002
[58]Ruud MBolle, Baba C Vemuri. On three-dimensional surface reconstruction. IEEE PAMI,1991,13(1):1-12
[59]仇芝.三维重建的匹配技术[D].南京理工大学,2004
[60]ILRIS-3D Operation Manual [M]. Optech Incorporated Industrial & 3D Imaging Divsion.2006
[61]赵艳平,高明等.基于PolyWorks的逆向工程数据处理[J].计算机应用技术,2005,32(10)：38-40
[62]许建中,马利庄.改进的Delaunay三角网渐次插入生成算法[J].计算机工程,2008,34(17)：254-256