车载激光点云的颜色信息获取与融合处理
|
摘要
车载激光扫描系统能提供实时、多尺度、多分辨率的数据,它的出现极大地满足了数字城市对数据采集效率的要求,是空间数据获取技术研究的热点。
车载激光扫描系统是多传感器集成的系统,主要由三大部分组成,分别是激光扫描仪、CCD相机、GPS/IMU.其中,激光扫描仪获取目标几何形状信息,CCD相机采集目标纹理信息,GPS/IMU得到车载系统位置与姿态。通过良好设计的空间同步和时间同步模式,可以在获取三维激光点云数据的同时获取相应的光学影像,融合处理激光点云和光学影像能实现城市场景的真彩色三维可视化,获得对城市场景的一致性解释与描述,展现了更优越的性能。颜色作为最直观、最感性的信息在其中有着举足轻重的作用。
本文主要围绕车载激光扫描系统中颜色数据的获取与融合处理进行研究,从整体出发讨论如何在车载系统的基础上对颜色的采集,颜色的处理进行改进。系统论述了颜色采集系统建立的原则,线阵CCD相机的工作原理,图像数据结构要求与格式转换,重点实验了基于直接线性变换和视平面与视线方程的线阵相机标定方法,着重阐述了建立颜色数据与激光扫描数据的映射关系的方法,并比较了基于最佳扫描行的融合方法与基于硬件的融合方法的优缺,分析了原始彩色激光点云的特点,提出了基于数值内插的方法来提升彩色激光点云的颜色纹理分辨率,从而增强城市场景真彩色三维可视化的效果,并对结果进行了分析。最后总结了颜色与点云交互处理的实例,强调颜色信息在点云应用上的巨大作用。
本文所实现的方法和程序都具有一定的实用性。无论是原始图像的转换,激光点云与纹理数据的融合,还是彩色点云的内插,基于颜色的点云应用,都能提供良好、稳定的结果,且充分考虑了对系统性能的需求。
Vehicle-borne Laser Scanning has an immense value in present days for fulfilling the demands of fast and cost-effective data acquisition for digital urban and other mapping related fields. It has flexibility in providing multi temporal, multi-dimensional and multi resolution data and has become central issue of spatial data acquisition.
The Vehicle-borne Laser Scanning of multiple sensors comprises of three basic components, viz.Laser Scanner, CCD camera and GPS/IMU. Laser Scanner provides range data (angle and distance from the sensor to each of the observed feature). CCD camera is used for image data and GPS/IMU are used for positioning. Through good space synchronization and time synchronization control, it can obtain 3D laser point clouds and corresponding optical images simultaneously. The true chromatic model of city scene may be realized with the help of integration of laser point clouds and optical images. Data fusing can explain and describe the consistency of urban scene, and show more superior performance. Texture being the most visible, most sensible has a pivotal role.
This paper focus on the color data acquisition and processing, discuss how to collect color and improve the visualization of the color from the overall based on available vehicle-borne laser scanning system. The platform and principle of color texture data obtaining are analyzed systematically. Line CCD camera principle of work is explained also. Image data structure and format transformation are defined too. Calibration of Line CCD Camera used Direct Linear Transformation method combined with the view plane and view line equation, because of its unique characteristics. As a key, the mapping model is built between the point clouds and digital images and as a result color point clouds is generated and compare the advantages and disadvantages of fusion based on best scan line between based on hardware. The characteristics of the original color point clouds are stated, numerical interpolation is proposed to ascend the resolution of color point clouds, in order to enhance true color 3D visualization's effect of city scene. Experiments are conducted and the results are clarified. Finally, the interactive applications of point clouds and color texture are summarized, the huge role of color in the applications of point cloud is emphasized.
The methods and programs implemented in this paper have practical value and can give good and stable results whether in initial images transformation、fusion between laser data and texture image and interpolation of initial color point clouds or applications of points cloud based on color texture. At the same time, system performance is fully considered.
车载激光扫描系统是多传感器集成的系统,主要由三大部分组成,分别是激光扫描仪、CCD相机、GPS/IMU.其中,激光扫描仪获取目标几何形状信息,CCD相机采集目标纹理信息,GPS/IMU得到车载系统位置与姿态。通过良好设计的空间同步和时间同步模式,可以在获取三维激光点云数据的同时获取相应的光学影像,融合处理激光点云和光学影像能实现城市场景的真彩色三维可视化,获得对城市场景的一致性解释与描述,展现了更优越的性能。颜色作为最直观、最感性的信息在其中有着举足轻重的作用。
本文主要围绕车载激光扫描系统中颜色数据的获取与融合处理进行研究,从整体出发讨论如何在车载系统的基础上对颜色的采集,颜色的处理进行改进。系统论述了颜色采集系统建立的原则,线阵CCD相机的工作原理,图像数据结构要求与格式转换,重点实验了基于直接线性变换和视平面与视线方程的线阵相机标定方法,着重阐述了建立颜色数据与激光扫描数据的映射关系的方法,并比较了基于最佳扫描行的融合方法与基于硬件的融合方法的优缺,分析了原始彩色激光点云的特点,提出了基于数值内插的方法来提升彩色激光点云的颜色纹理分辨率,从而增强城市场景真彩色三维可视化的效果,并对结果进行了分析。最后总结了颜色与点云交互处理的实例,强调颜色信息在点云应用上的巨大作用。
本文所实现的方法和程序都具有一定的实用性。无论是原始图像的转换,激光点云与纹理数据的融合,还是彩色点云的内插,基于颜色的点云应用,都能提供良好、稳定的结果,且充分考虑了对系统性能的需求。
Vehicle-borne Laser Scanning has an immense value in present days for fulfilling the demands of fast and cost-effective data acquisition for digital urban and other mapping related fields. It has flexibility in providing multi temporal, multi-dimensional and multi resolution data and has become central issue of spatial data acquisition.
The Vehicle-borne Laser Scanning of multiple sensors comprises of three basic components, viz.Laser Scanner, CCD camera and GPS/IMU. Laser Scanner provides range data (angle and distance from the sensor to each of the observed feature). CCD camera is used for image data and GPS/IMU are used for positioning. Through good space synchronization and time synchronization control, it can obtain 3D laser point clouds and corresponding optical images simultaneously. The true chromatic model of city scene may be realized with the help of integration of laser point clouds and optical images. Data fusing can explain and describe the consistency of urban scene, and show more superior performance. Texture being the most visible, most sensible has a pivotal role.
This paper focus on the color data acquisition and processing, discuss how to collect color and improve the visualization of the color from the overall based on available vehicle-borne laser scanning system. The platform and principle of color texture data obtaining are analyzed systematically. Line CCD camera principle of work is explained also. Image data structure and format transformation are defined too. Calibration of Line CCD Camera used Direct Linear Transformation method combined with the view plane and view line equation, because of its unique characteristics. As a key, the mapping model is built between the point clouds and digital images and as a result color point clouds is generated and compare the advantages and disadvantages of fusion based on best scan line between based on hardware. The characteristics of the original color point clouds are stated, numerical interpolation is proposed to ascend the resolution of color point clouds, in order to enhance true color 3D visualization's effect of city scene. Experiments are conducted and the results are clarified. Finally, the interactive applications of point clouds and color texture are summarized, the huge role of color in the applications of point cloud is emphasized.
The methods and programs implemented in this paper have practical value and can give good and stable results whether in initial images transformation、fusion between laser data and texture image and interpolation of initial color point clouds or applications of points cloud based on color texture. At the same time, system performance is fully considered.
引文
[1]Boyer W.S, Smith G.E., Bell System Technological journal,1970,49:587
[2]Ackermann. Airborne Laser Scanning-Present Status and Future Expectations[C]. ISPRS Journal of Photogrammetry and Remote Sensing,1999,54:64-67
[3]李树楷,刘少创.机载／车载激光扫描测距／成像制图系统的新进展[J].中国图形图象学报,1999,2(4)：89-94
[4]叶泽田,刘先林,钟若飞,刘百奇.车载多传感器集成快速数据获取与处理系统研究[J].中国科技成果,2008(12)：46-48
[5]陈永芳,叶泽田.基于多传感器融合的车载移动测图系统研究[J].测给通报,2007(1)：5-7
[6]陈永芳,叶泽田,张恒铁,石波.多传感器组合的移动车载数据采集系统研究[J].传感器与微系统,2006,25(12)：23-25
[7]Lynx Mobile Mapper, http://www.optech.ca/lynx.htm
[8]StreetMapper. http://streetmapper.net/index.htm
[9]Huijing Zhao, Ryosuke Shibasaki. Reconstructing textured CAD modeled of urban environment using vehicle-borne laxer range scanners and line cameras[A].International Workship on Computer Vision Systems Proceedings[C].Lecture Notes In Computer Science,2001.2095:284-297.
[10]Huijing Zhao, Ryosuke Shibasaki. Reconstruction of textured urban 3D model by fusing ground-based laser ranger and CCD images[J].IEICE Trans Inf & Syst,2000,E83-D(7):1429-1440
[11]张爱武,胡少兴,孙卫东,李风亭.基于激光与可见光同步数据的室外场景三维重建[J].电子学报,2005,33(5)：810-815.
[12]S.F.EI-Hakim, C.Brenner, G.Roth. A Multi-sensor Approach to Creating Accurate Virtual Environment[J]. JSPRS, Journal of Photogrammetry & Remote Sensing.1998, May(4): 113-117
[13]Schenk.T, Csath.B. Fusion of LIDAR data and serial imagery for a more completer surface description[J]. The International Archives of Photogrammetry and Remote Sensing and Spatial Information Sciences.2002,34(3A):310-317
[14]Rottensteiner, F.Trinder, J.Clode, S.Kubik, K. K. T. Fusing airborne laser scanner data and aerial imagery for the automatic extraction of buildings in densely built up areas[C]. Proceedings of the 2004th ISPRS Congress at Istanbul, The International Archives of Photogrammetry and Remote Sensing and Spatial Information Sciences.2004,512-517
[15]R.Reulke, A.Wehr. Mobile panoramic mapping using CCD-line camera and laser scanner with integrated position and orientation system[J]. In Proceedings of the ISPRS workshop group V/1.2004:165-183
[16]Peter Biberk, Henrik Andreasson.3D Modeling of Indoor Environments by a Mobile Robot with a Laser Scanner and Panoramic Camera[C], In International Symposium on Experimental Robotics(ISER),2005
[17]P.Ronnholm, E.Honkavaara, P.Litkey, H.Hyyppa, J.Hyyppa. INTEGRATION OF LASER SCANNING AND PHOTOGRAMMETRY[J].IAPRS Volume XXXVI, Part3/W52,2007
[18]徐胜华,朱庆.摄影测量三维重建中多源信息融合方法探讨[J].地理与地理信息科学,2005,6(21)：33-36
[19]张爱武,胡少兴,孙卫东,李风亭.基于激光与可见光同步数据的室外场景三维重建[J].电子学报,2005,33(5)：810-815
[20]李茂西,王从军,陈鑫.基于单目CCD摄影机的三维点云数据重建研究[J].自动化与仪表,2005(5)：13-16
[21]张栋.基于LIDAR数据和航空影像的城市房屋三维重建[D].武汉大学,2005
[22]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,2006,2(16)：297-302
[23]周克勤,赵煦,丁延辉.基于激光点云的三维可视化方法[J].测绘科学技术学报,2006,23(1)：69-72
[24]邓非,张祖勋,张剑清.一种激光扫描数据与数码照片的配准方法[J].武汉大学学报.(信息科学版),2007,4(32)：290-296
[25]张帆,黄先锋,李德仁.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
[26]李志林,朱庆.数字高程模型[M].武汉测绘科技大学出版社,2000=125-139
[27]戴雯.基于文物的表面纹理重建技术研究[D].江苏大学,硕士学位论文.2009.5
[28]丁立军.基于摄影原理的文物表面纹理重建技术研究[D].江苏大学,硕士学位论文.2007.3
[29]TVI-XIIMUS 4096 CT用户手册[M].2007.12
[30]杨长强,叶泽田,钟若飞.基于时空匹配的车载激光点云与CCD线阵图像的融合[J].测绘科学,2010,2(35)：32-33,182
[31]游文虎,姜复兴.INS/GPS组合导航系统的数据同步技术研究[J].中国惯性技术学报,2003,4(11)：20-22
[32]曲文乾.INS/DGPS支持的车载线阵CCD标定方法的研究及应用[D].北京：首都师范大学,2009
[33]冀翼IMU/DGPS辅助车载线阵CCD相机推扫路面图像的几何校正方法研究[D].北京：首都师范大学,2008
[34]李学友.IMU/DGPS辅助航空摄影综述[J].测绘科学,2005,5(30)：110-113
[35]卡尔曼滤波在INS/GPS组合导航中的应用研究[D].太原,中北大学,2006
[36]刘军,张永生,王冬红,徐卫明.POS AV510-DG系统外方位元素的计算方法[J].测绘技术装备,2004,4(6)：6-9
[37]朱述龙,史文中,张艳,朱宝山.线阵推扫式影像近似几何校正算法的精度比较[J].遥感学报,2005,3(8)：220-226
[38]张广军.视觉测量[M].北京：科学出版社,2008：27-35
[39]Abdel-Aziz,Yl.,Karara, H.M.Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry
[40]DLT (Direct Linear Transformation) Method, http://www.kwon3d.com/theory/dlt/dlt.html
[41]Radu Horaud,Roger Mohr, Boguslaw Lorecki. On Single-Scanline Camera Calibration[J]. IEEE TRANSACTIONS ON ROBOBTICS AND AUTOMATION.1993,9(1):71-75
[42]Yihong Wu,Youfu Li,Zhanyi Hu.Detecting and Handling Unreliable Points for Camera Parameter Estimation[J]. International Journal of Computer Vision,2008,79:209-223
[43]刘军,张永红,王冬生,徐卫明INS/DGPS支持的机型线阵推扫影像几何校正[J].遥感学 报,2006,1(10)：21-26
[44]张浩,张兵,卫征.基于POS的高光谱影像快速几何纠正[J].测绘通报,2009,1：14-17
[45]杨勇.车载三维激光点云数据与单目CCD相机纹理影像数据的融合[D].北京,首都师范大学,2010
[46]D Manandhar, R Shibasaki. VEHICLE-BORNE LASER MAPPING SYSTEM(VLMS)FOR 3-D URBAN GIS DATABASE[C]. IEEE Geoscience and Remote Sensing Symposium,2001,C(00): 2073-2075
[47]贾永红.数字图像处理[M].武汉大学出版社,2001
[48]李新,程国栋,卢玲.空间内插方法比较[J].地球科学进展,2000,3(15)：260-265
[49]杨必胜,魏征,李清泉,毛庆洲.面向车载激光扫描点云快速分类的点云特征图像生成方法[J].测绘学报,2010,5(30)：540-545
[50]Susanne Becker, Norbert Haala. COMBINED FEATURE EXTRACTION FOR FACADE RECONSTRUCTION[C].ISPRS Workshop on Laser Scanning 2007 and SilviLaser 2007, Espoo: ISPRS,2007:44-49
[51]UMEDA K, IKUSHIMA K, ARAI T 3D Shape Recognition by Distributed Sensing of Range Images and Intensity lmages[C]. Proceedings of the 1997 IEEE International Conference on Robotics and Automation. Albuquerque:IEEE,1997:149-154
[2]Ackermann. Airborne Laser Scanning-Present Status and Future Expectations[C]. ISPRS Journal of Photogrammetry and Remote Sensing,1999,54:64-67
[3]李树楷,刘少创.机载／车载激光扫描测距／成像制图系统的新进展[J].中国图形图象学报,1999,2(4)：89-94
[4]叶泽田,刘先林,钟若飞,刘百奇.车载多传感器集成快速数据获取与处理系统研究[J].中国科技成果,2008(12)：46-48
[5]陈永芳,叶泽田.基于多传感器融合的车载移动测图系统研究[J].测给通报,2007(1)：5-7
[6]陈永芳,叶泽田,张恒铁,石波.多传感器组合的移动车载数据采集系统研究[J].传感器与微系统,2006,25(12)：23-25
[7]Lynx Mobile Mapper, http://www.optech.ca/lynx.htm
[8]StreetMapper. http://streetmapper.net/index.htm
[9]Huijing Zhao, Ryosuke Shibasaki. Reconstructing textured CAD modeled of urban environment using vehicle-borne laxer range scanners and line cameras[A].International Workship on Computer Vision Systems Proceedings[C].Lecture Notes In Computer Science,2001.2095:284-297.
[10]Huijing Zhao, Ryosuke Shibasaki. Reconstruction of textured urban 3D model by fusing ground-based laser ranger and CCD images[J].IEICE Trans Inf & Syst,2000,E83-D(7):1429-1440
[11]张爱武,胡少兴,孙卫东,李风亭.基于激光与可见光同步数据的室外场景三维重建[J].电子学报,2005,33(5)：810-815.
[12]S.F.EI-Hakim, C.Brenner, G.Roth. A Multi-sensor Approach to Creating Accurate Virtual Environment[J]. JSPRS, Journal of Photogrammetry & Remote Sensing.1998, May(4): 113-117
[13]Schenk.T, Csath.B. Fusion of LIDAR data and serial imagery for a more completer surface description[J]. The International Archives of Photogrammetry and Remote Sensing and Spatial Information Sciences.2002,34(3A):310-317
[14]Rottensteiner, F.Trinder, J.Clode, S.Kubik, K. K. T. Fusing airborne laser scanner data and aerial imagery for the automatic extraction of buildings in densely built up areas[C]. Proceedings of the 2004th ISPRS Congress at Istanbul, The International Archives of Photogrammetry and Remote Sensing and Spatial Information Sciences.2004,512-517
[15]R.Reulke, A.Wehr. Mobile panoramic mapping using CCD-line camera and laser scanner with integrated position and orientation system[J]. In Proceedings of the ISPRS workshop group V/1.2004:165-183
[16]Peter Biberk, Henrik Andreasson.3D Modeling of Indoor Environments by a Mobile Robot with a Laser Scanner and Panoramic Camera[C], In International Symposium on Experimental Robotics(ISER),2005
[17]P.Ronnholm, E.Honkavaara, P.Litkey, H.Hyyppa, J.Hyyppa. INTEGRATION OF LASER SCANNING AND PHOTOGRAMMETRY[J].IAPRS Volume XXXVI, Part3/W52,2007
[18]徐胜华,朱庆.摄影测量三维重建中多源信息融合方法探讨[J].地理与地理信息科学,2005,6(21)：33-36
[19]张爱武,胡少兴,孙卫东,李风亭.基于激光与可见光同步数据的室外场景三维重建[J].电子学报,2005,33(5)：810-815
[20]李茂西,王从军,陈鑫.基于单目CCD摄影机的三维点云数据重建研究[J].自动化与仪表,2005(5)：13-16
[21]张栋.基于LIDAR数据和航空影像的城市房屋三维重建[D].武汉大学,2005
[22]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,2006,2(16)：297-302
[23]周克勤,赵煦,丁延辉.基于激光点云的三维可视化方法[J].测绘科学技术学报,2006,23(1)：69-72
[24]邓非,张祖勋,张剑清.一种激光扫描数据与数码照片的配准方法[J].武汉大学学报.(信息科学版),2007,4(32)：290-296
[25]张帆,黄先锋,李德仁.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
[26]李志林,朱庆.数字高程模型[M].武汉测绘科技大学出版社,2000=125-139
[27]戴雯.基于文物的表面纹理重建技术研究[D].江苏大学,硕士学位论文.2009.5
[28]丁立军.基于摄影原理的文物表面纹理重建技术研究[D].江苏大学,硕士学位论文.2007.3
[29]TVI-XIIMUS 4096 CT用户手册[M].2007.12
[30]杨长强,叶泽田,钟若飞.基于时空匹配的车载激光点云与CCD线阵图像的融合[J].测绘科学,2010,2(35)：32-33,182
[31]游文虎,姜复兴.INS/GPS组合导航系统的数据同步技术研究[J].中国惯性技术学报,2003,4(11)：20-22
[32]曲文乾.INS/DGPS支持的车载线阵CCD标定方法的研究及应用[D].北京：首都师范大学,2009
[33]冀翼IMU/DGPS辅助车载线阵CCD相机推扫路面图像的几何校正方法研究[D].北京：首都师范大学,2008
[34]李学友.IMU/DGPS辅助航空摄影综述[J].测绘科学,2005,5(30)：110-113
[35]卡尔曼滤波在INS/GPS组合导航中的应用研究[D].太原,中北大学,2006
[36]刘军,张永生,王冬红,徐卫明.POS AV510-DG系统外方位元素的计算方法[J].测绘技术装备,2004,4(6)：6-9
[37]朱述龙,史文中,张艳,朱宝山.线阵推扫式影像近似几何校正算法的精度比较[J].遥感学报,2005,3(8)：220-226
[38]张广军.视觉测量[M].北京：科学出版社,2008：27-35
[39]Abdel-Aziz,Yl.,Karara, H.M.Direct linear transformation from comparator coordinates into object space coordinates in close-range photogrammetry
[40]DLT (Direct Linear Transformation) Method, http://www.kwon3d.com/theory/dlt/dlt.html
[41]Radu Horaud,Roger Mohr, Boguslaw Lorecki. On Single-Scanline Camera Calibration[J]. IEEE TRANSACTIONS ON ROBOBTICS AND AUTOMATION.1993,9(1):71-75
[42]Yihong Wu,Youfu Li,Zhanyi Hu.Detecting and Handling Unreliable Points for Camera Parameter Estimation[J]. International Journal of Computer Vision,2008,79:209-223
[43]刘军,张永红,王冬生,徐卫明INS/DGPS支持的机型线阵推扫影像几何校正[J].遥感学 报,2006,1(10)：21-26
[44]张浩,张兵,卫征.基于POS的高光谱影像快速几何纠正[J].测绘通报,2009,1：14-17
[45]杨勇.车载三维激光点云数据与单目CCD相机纹理影像数据的融合[D].北京,首都师范大学,2010
[46]D Manandhar, R Shibasaki. VEHICLE-BORNE LASER MAPPING SYSTEM(VLMS)FOR 3-D URBAN GIS DATABASE[C]. IEEE Geoscience and Remote Sensing Symposium,2001,C(00): 2073-2075
[47]贾永红.数字图像处理[M].武汉大学出版社,2001
[48]李新,程国栋,卢玲.空间内插方法比较[J].地球科学进展,2000,3(15)：260-265
[49]杨必胜,魏征,李清泉,毛庆洲.面向车载激光扫描点云快速分类的点云特征图像生成方法[J].测绘学报,2010,5(30)：540-545
[50]Susanne Becker, Norbert Haala. COMBINED FEATURE EXTRACTION FOR FACADE RECONSTRUCTION[C].ISPRS Workshop on Laser Scanning 2007 and SilviLaser 2007, Espoo: ISPRS,2007:44-49
[51]UMEDA K, IKUSHIMA K, ARAI T 3D Shape Recognition by Distributed Sensing of Range Images and Intensity lmages[C]. Proceedings of the 1997 IEEE International Conference on Robotics and Automation. Albuquerque:IEEE,1997:149-154