摘要
对于工作在非结构化场景中的移动机器人系统,能否对其周围环境进行有效辨识、感知和重构是其能够自主运行的前提条件。尤其对三维非结构化场景的建模与重构的研究,具有更重要的研究意义和广泛的应用价值。本文主要完成了三维场景重构系统的构建,有效建立了场景的几何特征模型和三维栅格模型,实现了对三维场景的重构与回溯等研究工作。
本文利用三维激光扫描数据的有序性,快速提取出非结构化场景中的直线、平面以及边缘轮廓等几何特征,建立三维场景的几何特征模型。为了实现大范围场景数据的调度及可视化,基于八叉树空间划分的思想,建立场景的三维栅格模型,将空间中的激光数据划分到三维栅格中。
经典ICP算法通过查找最近点构成匹配对,以迭代的方式实现场景匹配。该方法具有迭代速度慢,且易收敛到局部极小值等缺点。本文提出了使用边缘特征点代替原始点云等五点改进策略,有效地克服了经典ICP算法的不足。为解决ICP算法收敛到局部极小值的缺点,需要利用扫描视点位姿进行初始校正。然而,扫描视点位姿往往不能有效提供,且存在较大误差。因此本文提出利用视觉图像处理的方法,辅助三维激光场景的匹配,从而实现无位姿的场景匹配问题。将视觉摄像机与三维激光测距仪结合,通过标定的方法标定出他们之间参数关系,得到视觉图像与三维激光深度图像之问的映射。首先对两幅场景对应的视觉图像进行处理,利用SIFT算法查找两幅图像中的特征匹配对,再将特征匹配对映射到激光场景中,构成三维激光特征匹配对,最后计算出两幅场景的转换关系,实现无位姿场景匹配。
基于三维栅格模型,本文解决了大范围场景的数据分块及实时调度显示的问题。根据OpenGL空间中的虚拟摄像机当前位置,按离摄像机的远近分5个显示等级,实时调度并显示栅格模型中相应的数据。通过实验证明,本文方法能够有效实现非结构化场景的建模与重构,具有很强的实用性。
Effective recognition and perception of environment are two essential tasks for mobile robot's navigation and exploration while working in the unstructured scenes. Especially for 3D scenes modeling and reconstruction, the study has a wide range of applications. The intelligent equipment for 3D scene reconstruction is built in this thesis. Moreover, the geometrical model and 3D grid model are built for 3D scenes. This thesis completes the reconstruction and back review of the 3D scenes.
According to the orderliness of laser scanning, the geometrical model is built by extracting the lines, planes, edge contours etc. To complete the data dispatching and visualization in large-scale scenes, a 3D grid model is built by using the octree space partition idea, which divides the laser data into small 3D grids effectively.
The classical ICP algorithm completes scene matching iteratively by finding the nearest points to constitute matching pairs. This approach has the disadvantages of slow iterating and converges to a local minimum easily. In this thesis we propose five ways to improve these disadvantages. The scanning pose is needed while matching scenes using ICP. Unfortunately, the scanning pose can not be provided effectively in the most time. Hence, a new method is proposed in this thesis to solve this problem. In the research, a camera is used together with the 3D laser scanner. The mapping relationship of the digital image and the laser range image is obtained by calibrating the parameters between the two kinds of sensors. After the matching pairs in the corresponding digital images are found by the SIFT algorithm, we map them to the laser range image to get the 3D laser matching pairs. At last, the transformation of the two scenes is calculated without scanning pose.
Based on the 3D grid model, the data blocking and dispatching of large-scale scenes are solved. According to the distance of grid to the virtual camera in OpenGL space, we divide five areas to dispatch the laser data in 3D grid model. The nearer to the virtual camera, the more data is dispatched and displayed on the screen. A series of experiment results and data analysis show the method's validity and practicability.
引文
[1] DEBEVEC P E, TAYLOR C J, MALIK J. Modeling and rendering architecture from photographs: A hybrid geometry and image-based approach [C]. In SIGGRAPH' 96, 1996:11-20.
[2] SEQUEIRA V, GONCALVES J, RIBEIRO M. 3D environment modelling using laser range sensing [J]. Robotics and Automation 1995,16:81-91.
[3] SEQUEIRA V, NG K, WOLFART E, et al. Automated 3D reconstruction of interiors with multiple scan - views [C]. In Proceedings of SPIE, Electronic Imaging' 99, The Society for Imaging Science and Technology /SPIE' s 11th Annual Symposium, 1999, 1.
[4] El-HAKIM S F, BOULANGER P, BLAIS F, et al. A system for indoor 3D mapping and virtual environments [C]. SPIE Proc Videometrics, 1997, 3174:21-35.
[5] El-HAKIM S F, BRENNER C, ROTH G. A multi-sensor approach to creating accurate virtual environment [J]. Journal of Photo grammetry & Remote Sensing. 1998:379-391.
[6] STAMOS I, LEORDEANU M. Automated Feature-Based Registration of Urban Scenes of Large Scale [C]. Proceeding of the 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003:555-561.
[7] STAMOS I, ALLEN P K. Geometry and texture recovery of scenes of large scale [J]. Computer Vision and Image Understanding. 2002, 11:94-118.
[8] STAMOS I, ALLEN P K. Registration of 3D with 2D imagery in urban environments [C].International Conference on Computer Vision. 2001,7:731-736.
[9] HAHNEL D, BURGARD W, THRUN S. Learning Compact 3D Models of Indoor and Outdoor Environments with a Mobile Robot [J]. Robotics and Autonomous Systems, 2003,44(1):15-27.
[10] THURUN S, FOX D, BURGARD W. A real-time algorithm for mobile robot mapping with application to multi robot and 3D mapping [C]. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA ' 00), San Francisco, CA, USA, 2000,4.
[11] WULF O, ARRAS K O, CHRISTENSEN H I, et al. 2D Mapping of Cluttered Indoor Environments by Means of 3D Perception [C]. In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA ' 04), New Orleans, USA, 2004,4:4204-4209. [12] SURMANN H, LINGEMANN K, NUCHTER A, et al. Fast acquiring and analysis of three dimensional laser range data [C]. Proceedings of the Sixth International Fall Workshop Vision, Modeling and Visualization, 2001,11:59-66.
[13] SURMANN H, NUCHTER A, HERTZBERG J. An autonomous mobile robot with a 3D laser range finder for 3D exploration and digitalization of indoor environments [J]. Robotics and Autonomous system. 2003,45:181-198.
[14]SURMANN H,NUCHTER A,LINGEMANN K,et al.6D SLAM A Preliminary Report on Closing the Loop in Six Dimensions[C].In Proc.IFAC Symp,IAV,Lisbon,Portugal,2004.
[15]NUCHTER A,LINGEMANN K,HERTZBERG J,et al.Mapping of Rescue Environments with Kurt3D [C].Proceedings of the International Workshop on Safety,Security and Rescue Robotics (SSRR '05),2005:158-163.
[16]NUCHTER A,LINGEMANN K,and HERTZBERG J.Kurt3D - A Mobile Robot for 39 Mapping of Environments[P],ELROB Technical Paper,Hammelburg,Germany,2006,5.
[17]NUCHTER A,LINGEMANN K,and HERTZBERG J,et al.69 SLAM-3D Mapping Outdoor Environments,Journal of Field Robotics(JFR),Special Issue on Quantitative Performance Evaluation of Robotic and Intelligent Systems,2007,24(8-9):699-722.
[18]BIBER P,ANDREASSON H,DUCKETT T,et al.3D Modeling of Indoor Environments by a Mobile Robot with a Laser Scanner and Panoramic Camera[J].In Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems(IROS' 04),Sendai,Japan.
[19]OHNO K,TADOKORO S.Dense 3D map building based on LRF data and color image fusion [C].In IEEE Int.Conf.on Intelligent Robots and Systems,2005:2792-2797.
[20]HU S,ZHAH,ZHANG A.Registration of Multiple Laser Scans Based on 3D Contour Features,10th Int.Conf.on Information Visualization,725-730,London,UK,2006,7.
[21]杨明,王宏,何克忠,等.基于激光雷达的移动机器人环境建模与避障[J],清华大学学报(自然科学版),2000,40(7):112-116.
[22]MILLER I,CAMPBELL M.A Mixture-Model Based Algorithm for Real-Time Terrain Estimation [J].Journal of Robotic Systems,September,2006,23(9):755-775.
[23]BESL P and MAKAY N.A method for Registration of 3D Shapes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1992,14(2):239-256.
[24]LOWED G.Object recognition from local scale-invariant features[C],International Conference on Computer Vision,Corfu,Greece,1999:1150-1157.
[25]LOWE D G.Distinctive image features from scale-invariant keypoints[J],International Journal of Computer Vision,2004,60(2):91-110.
[26]SICK AG Auto Ident.Technical description LMS 200 laser measurement system[P].Germany:SICK AG Auto Ident,2000.
[27]ZHANG Z.A flexible new technique for camera calibration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(11):1330-1334.
[28]SCARAMUZZA D,HARATI A,SIEGWART R.Extrinsic Self Calibration of a Camera and a 3D Laser Range Finderfrom Natural Scenes.IEEE,2007:4164-4169.
[29]和平鸽工作室.OpenGL三维图形系统开发与使用技术[M].北京:清华大学出版社,2003.
[30]李志华,徐静,张庆顺,等.利用OpenGL实现屏幕坐标与世界坐标的相互变换,河北建筑科技学院学报[J],2005,22(3):83-86.
[31]顾露.使用OpenGL实现三维坐标的鼠标拣选[J],软件工程,2004,4:21-25.
[32]FREY P J,BOROUCHAKI H,GEORGE P L.3D Delauny mesh generation coupled with an advancing front approach[J],compute Methods Appl.Mech.Engrg.1998,157:115-131.
[33]WEATHERILL N,HASSAN O.Efficient three dimension Delauny triangulation with automatic point creation and imposed boundary constrains[J],Int.J.Numer.Methods Engrg.1994,37:2005-2039.
[34]COLEMAN S A,SCOTNEY B W,SUGANTHAN S.Feature Extraction on Range Images -A New Approach [C].Proceedings of IEEE International Conference on Robotics and Automation,Rome,2007:1098-1103.
[35]FRANKLIN D,FIRBY R].Integrating Range and Object Data for Robot Navigation[C].In Proceedings of the first international conference on Autonomous agents,Marina del Rey,California,United States,1997:185-192.
[36]GONZALEZ J,OLLERO A,REINA A.Map Building for a Mobile Robot equipped with a 2D Laser Rangefinder[C],IEEE Int.Conf.Robotics and Automation,1994,3:1904-1909.
[37]孙殿柱,范志先,李延瑞,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):23-45.
[38]路兴昌,宫辉力,赵文吉,等.基于激光扫描数据的三维可视化建模[J],系统仿真学报,2007,7(19):53-60.
[39]黄国彦.结合三维网格及迭代式Hough变换进行光达点云资料结构化处理[D],国立台湾大学土木所论文,2005.
[40]曾义星,刘嘉铭.光达点云空间特征之分类、定义及萃取方法[C],台湾“内政部‘辩理LIIDAR 测区之高精度及高解析度数值地形测绘、资料库建置与应用推广工作案'成果发表暨应用研讨会”,2005,12.
[41]HUJAZI A E.Range Image Segmentation with applications to Robot Bin-Picking Using Vacuum Gripper[J].IEEE Trans.Systems,Man,and Cybernetics,1990,20(6):1313-1325.
[42]JIANG X Y,BUNKE H.Edge detection in range image based on scan line approximation.Computer Vision ad Image Understanding,1999,73(2):183-199.
[43]KRISHNAPURAM R,GUPTA S.Morphological Methods for Detection and Classification for Edges in Range Images[J].Journal of Mathematical Imaging Vision,1992:351-375.
[44]SUGANTHAN S,COLEMAN S A,SCOTNEY B W.Scalable Operators for Feature Extraction on 3-D Data[J].Springer Tracts in Advanced Robotics 2008,44:263-272.
[45]SCHNABEL R,KLEIN R.Octree Based Point Cloud Data Compression In Proceedings of the Eurographics Symposium on Point Based Graphics,2006:15-23.
[46]吴晓军,刘伟军,王天然.基于八叉树的三维网格模型体素化方法[J],工程图学学报,2005,4:15-22.
[47]叶至军.C++STL开发技术导引[M].北京:人民邮电出版社,2007.
[48]刘晓利,彭翔,殷永凯,等.借助标志点的深度数据全局匹配方法[J].光学学报,2009,29(4):1010-1014.
[49]LORUSSO A,EGGERT D,FISHER R.A Comparison of Four Algorithms for Estimating 3-D Rigid Transformations[C].In:Proceedings of the 4th British Machine Vision Conference(BMVC 1995),Birmingham,England,1995:237-246.
[50]ARUN K S,HUANG T S,BLOSTEIN S D.Least square fitting of two 3-d point sets[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1987,9(5):698-700.
[51]HORN B K P.Closed-form solution of absolute orientation using unit quaternions[J].Journal of the Optical Society of America,1987,4(4):629-642.
[52]孙宇,项志宇,刘济林.未知室外环境下移动机器人的三维场景重建[J].浙江大学学报(工学版),2007,41(12):1949-1954.
[53]MASUDA T,YOKOYA N.A Robust Method for Registration and Segmentation of Multiple Range Images.Computer Vision and Image Understanding.1995,61(3):295-307.
[54]DORAI C,WENG J,JAIN A.Registration and Integration of Multiple Object Views for 30 Model Construction[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.1998,20:83-89.
[55]GREENSPAN M,YURICK M.Approximate K-D Tree Search for Efficient ICP[J].In Proceedings of the 4th IEEE International Conference on Recent Advances in 30 Digital Imaging and Modeling(3DIM 2003),2003:442-448.
[56]FISHCHLER M A.Random sample consensus:a paradigm for model fitting with application to image analysis and automated cartography[J].Communication Association Machine,1981,24(6):381-395.
[57]BERGEVIN R,SOUCY M,GAGNON H,et al.Towards a general multi-view registration technique[J].IEEE Transactions on Pattern Analysis and Machine Intelligence.1996,18(5):540-547.
