基于无人机影像空中三角测量的研究
|
摘要
随着我国城镇化建设快速发展,城镇建设规模扩大,发展迅速，对遥感数据的现势性和及时性获取的要求越来越高，而无人机航空摄影作业方式灵活快捷、其获取的影像具有时效性、分辨率高等优点，因此无人机航空摄影在各行各业得到了越来越广泛的应用。但其获取的影像存在畸变差大、像幅小、数量多、基线短、重叠度不规则且倾角过大等问题，这些问题都是国内外现有的数字摄影测量系统面临的新问题。对于以上的这些问题，是如何通过算法进行解决处理的，对于算法解决不了的，实际生产中又是通过哪些生产方案来进行解决的，这些都让基于无人机航空摄影影像的空三加密的无论在理论上还是实际生产中都有一定的研究价值。
本文将以无人机航空摄影数据为研究对象，围绕无人机影像数据的特点、高精度相机标定、影像匹配、平差等几个方面对空中三角测量的理论知识进行展开，然后通过实际的生产试验来说明空中三角测量在实际生产的应用以及实际的生产方案，本文从以下几个方面展开全文：
（1）由于UAV系统使用的数码相机属于非量测型相机，其畸变差对实际数字产品影响较大。本文简单介绍其概念以及其改正模型，并说明非量测相机畸变差改正的必要性，通过实验说明了畸变纠正对空三精度的影响。
（2）由于UAV遥感系统平台的平稳程度不高，造成影像重叠度不规则、旋偏角大等原因，目前灰度匹配算法难以满足无人机影像匹配的要求。本文从影像匹配基本原理着手，介绍了目前常用的并且适用于无人机影像匹配算法，通过实验研究分析了基于Harris算子和SIFT特征匹配的实现过程和匹配效果。
（3）本文采用实际的海南无人机生产数据，对空三加密实际生产进行了试验研究与分析，详细提出匹配出一种特殊粗差点以及对该粗差点的剔除方式，最后对试验进行总结。
以上研究表明，对无人机影像进行畸变差改正是必不可少的，文中提及的基于Harris算子和SIFT特征的匹配能有效解决旋偏角大、亮度不一致、重叠不度不规则的无人机影像匹配的问题，实际应用研究发现，该匹配算法在大量匹配特征点时，会匹配出一类光束法无法剔除的粗差点，文中提出了从生产方案的角度进行剔除。
With the fast development and the scale expanding of the urbanization construction, so the demand for the current potential and timeliness of access to remote sensing data become higher and higher. Because UAV Aerial Photography have many advantages, for example, flexible and efficient practices, the images obtained with the timeliness and High resolution, so UAV aerial photography has been more widely used in all walks of life. However, the images obtained have such problems, for big Distortion、like pieces of small、the Quantity images、Baseline short、Degree of overlap irregular and Rotation angle is too large. These issues are new problems for the existing digital photogrammetric system facing at home and abroad. For the above problems, How be resolved by the algorithm to deal with. For the algorithm cannot be resolved, is how to resolve at the Actual production. These allow the research of aerial triangulation based on the UAV images have some value in terms of theory and actual production.
This will be UAV aerial photography data for the study, expand the article from the theoretical knowledge of aerial triangulation around the characteristics of the UAV image data, high-precision camera calibration, image matching, adjustment and other aspects. Then by the production experiments to illustrate Aerial triangulation in the actual production, this paper has researched mainly on the following work:
(1)Because used digital cameras in UAV system is a non-metric camera, the distortion difference has greater impact on the actual number of products. This paper briefly describes the concept as well as the correction mode, and explain the necessity of distortion error correction for non-metric camera. By the Experiments show the distortion correction affect the accuracy of aerial triangulation.
(2)Because UAV remote sensing platform is not stable, resulting the image overlay degree irregular, rotation angle large, so gray matching algorithm is difficult to meet the requirements of the UAV image matching. This article describes the basic principles of the image matching, Describes the common and applicable algorithm to the UAV image matching. By Experiment study the matching process and matching effect based on Harris and SIFT matching method.
(3)This paper Using Hainan UAV production data, introduces the research and analysis the Application of aerial triangulation in the actual production. This paper introduces the special error point due to matching, and introduces the way of removing the error point. Finally, summarizing the experiment.
The above studies show that, distortion correction for the UAV image is essential. The Harris and the SIFT matching algorithm mentioned in the paper can effectively solve the matching of UAV images. But the experiment finds most error points that can't be removed by bundle adjustment, the paper introduces a way of removing by production program.
本文将以无人机航空摄影数据为研究对象，围绕无人机影像数据的特点、高精度相机标定、影像匹配、平差等几个方面对空中三角测量的理论知识进行展开，然后通过实际的生产试验来说明空中三角测量在实际生产的应用以及实际的生产方案，本文从以下几个方面展开全文：
（1）由于UAV系统使用的数码相机属于非量测型相机，其畸变差对实际数字产品影响较大。本文简单介绍其概念以及其改正模型，并说明非量测相机畸变差改正的必要性，通过实验说明了畸变纠正对空三精度的影响。
（2）由于UAV遥感系统平台的平稳程度不高，造成影像重叠度不规则、旋偏角大等原因，目前灰度匹配算法难以满足无人机影像匹配的要求。本文从影像匹配基本原理着手，介绍了目前常用的并且适用于无人机影像匹配算法，通过实验研究分析了基于Harris算子和SIFT特征匹配的实现过程和匹配效果。
（3）本文采用实际的海南无人机生产数据，对空三加密实际生产进行了试验研究与分析，详细提出匹配出一种特殊粗差点以及对该粗差点的剔除方式，最后对试验进行总结。
以上研究表明，对无人机影像进行畸变差改正是必不可少的，文中提及的基于Harris算子和SIFT特征的匹配能有效解决旋偏角大、亮度不一致、重叠不度不规则的无人机影像匹配的问题，实际应用研究发现，该匹配算法在大量匹配特征点时，会匹配出一类光束法无法剔除的粗差点，文中提出了从生产方案的角度进行剔除。
With the fast development and the scale expanding of the urbanization construction, so the demand for the current potential and timeliness of access to remote sensing data become higher and higher. Because UAV Aerial Photography have many advantages, for example, flexible and efficient practices, the images obtained with the timeliness and High resolution, so UAV aerial photography has been more widely used in all walks of life. However, the images obtained have such problems, for big Distortion、like pieces of small、the Quantity images、Baseline short、Degree of overlap irregular and Rotation angle is too large. These issues are new problems for the existing digital photogrammetric system facing at home and abroad. For the above problems, How be resolved by the algorithm to deal with. For the algorithm cannot be resolved, is how to resolve at the Actual production. These allow the research of aerial triangulation based on the UAV images have some value in terms of theory and actual production.
This will be UAV aerial photography data for the study, expand the article from the theoretical knowledge of aerial triangulation around the characteristics of the UAV image data, high-precision camera calibration, image matching, adjustment and other aspects. Then by the production experiments to illustrate Aerial triangulation in the actual production, this paper has researched mainly on the following work:
(1)Because used digital cameras in UAV system is a non-metric camera, the distortion difference has greater impact on the actual number of products. This paper briefly describes the concept as well as the correction mode, and explain the necessity of distortion error correction for non-metric camera. By the Experiments show the distortion correction affect the accuracy of aerial triangulation.
(2)Because UAV remote sensing platform is not stable, resulting the image overlay degree irregular, rotation angle large, so gray matching algorithm is difficult to meet the requirements of the UAV image matching. This article describes the basic principles of the image matching, Describes the common and applicable algorithm to the UAV image matching. By Experiment study the matching process and matching effect based on Harris and SIFT matching method.
(3)This paper Using Hainan UAV production data, introduces the research and analysis the Application of aerial triangulation in the actual production. This paper introduces the special error point due to matching, and introduces the way of removing the error point. Finally, summarizing the experiment.
The above studies show that, distortion correction for the UAV image is essential. The Harris and the SIFT matching algorithm mentioned in the paper can effectively solve the matching of UAV images. But the experiment finds most error points that can't be removed by bundle adjustment, the paper introduces a way of removing by production program.
引文
[1]王琳.高精度、高可靠的无人机影像全自动相对定向及模型连接研究[J].2011.
[2]孙杰,林宗坚,崔红霞.无人机低空遥感监测系统[J].遥感信息,2003,(1)：4950,27.
[3]Sang-Jong Lee, Seong-pil Kim, Tae-Sik Kim, etal. Development of Autonomous Flight Control System for50m Unmanned Airship [J]. IS SNIP,2004,457-462.
[4]Filoktimon Repoulias, Evangelos Papadopoulos. Dynamically Feasible Trajectory and Open-Loop Control Design for Unmanned Airships[C].2007Mediterranean Conference on Control and Automation,2007.
[5]Jose Raul Azinheira, Ely Carneiro de Paiva, Josue Jr.G.Ramos, et al. Mission Path Following for an Autonomous Unmanned Airship [C]. International Conference on Robotics&Automation,2000.
[6]D.Gray, M.Nagatsuka, R.Miura. Autopsy of Unmanned Airship Television Broadcast [J]. Yokosuka Radio Communications Research Centre,2006.
[7]Herwitz R.S., Lee F.J., Arvesen J.C. Precision Agriculture as a Commercial Application for Solar-Powered Unmanned Aerial Vehicles [C].1st American Institue of Aeronautics and Astronautics UAV Conference,2002.
[8]Jinjun Rao, Zhenbang Gong, Jun Luo, et al. Unmanned Airships for Emergency Management[C]. International Workshop on Safety, Security and Rescue Robotics,2005.
[9]Dugan J. and Piotrowski C. Develpomental system for maritime rapid environment assessment using UAVs[C].5st ONERA_DLR aerospace symposium,2003.
[10]Robert M.N. Unmanned air vehicles-coming of age at last[J]. Air&space Europe,2000,2(1):23-25.
[11]Toth C.K, G-Brzezinska D., Merry C. Supporting traffic flow management with high-definition imagery[C]. ISPRS workshop on high resolution mapping from space2003. Hannover,2003.
[12]戴斌.航空摄影测量在土地利用现状更新调查中的应用[J].铁道勘察,2005,(6)：24-26.
[13]申海建,郭荣中,黄小波等.微型无人机(MUAV)航空摄影测量技术在土地整理项目规划设计中的应用[C].2007年中国土地学会年会,2007.
[14]张雪萍,刘英.无人机在大比例尺DOM生产中的应用[J].测绘标准化,2011,27(4)：25-27.
[15]李雷,陈桂琴,田园.无人机的发展历程与展望[J].科技论坛.
[16]http://chzt.sbsm.gov.cn/article/qt/xjsxcpcxyy/ckxt/jsjs/.新技术新产品创新应用.
[17]张静.“光照度平方反比定律”的实验设计[J].物理教学探讨,2009,(19)
[18]CJJ899,城市测量规范[S].
[19]林君建,苍桂华.摄影测量学[M].北京：国防工业出版社,2005.
[20]王青山.简述无人机在遥感技术中的应用[J],测绘与空间地理信息,2010,(3),101.
[21]马瑞升,孙涵,林宗桂,等.微型无人机遥感影像的纠偏与定位[J].南京气象学院学报,2005,28(5)：632-639.
[22]王俊.无人机航空摄影的空三评价分析[J].甘肃科技,2011,13(27)：40-43.
[23]宫阿都,何孝莹,雷添杰,李京.无控制点数据的无人机影像快速处理[J].地球信息科学学报,2010,2(12)：254-259.
[24]程效军,胡敏捷.数码相机畸变差的检测[J].测绘学报,2002,(31)：113-117.
[25]张祖勋,张剑清.数字摄影测量学.武汉：武汉测绘科技大学出版社,1997.216-232.
[26]宣家斌.航摄仪内方位元素和物镜畸变差的测定
[27]邹勇平,徐水平,李月锋,林沂才.实验室检定航摄相机内方位元素和畸变差计算方法的比较[J].测绘技术装备,2006,1：37-38.
[28]崔红霞,孙杰,林宗坚,储美华.非量测数码相机的畸变差检测研究[J].测绘科学,2005,1(30)：105-107.
[29]金仲华,崔红霞.数码相机径向畸变差检测及改正[J].科技广场,2008.
[30]冯文灏,商浩亮.成像系统通用数字畸变模型的建立原理与应用.数字近景摄影测量会议,武汉,2004.
[31]C.J.Harris, M.StePhens.A combined Corner and Edge Detector[C]. Proceedings of the Fourth Alvey Vision Conference,1988:147-151.
[32]王玉宏,陈永立.低空遥感影像匹配与拼接技术研究.测绘信息与工程,2009,34(2).
[33]TOET A, RUYVEN L, VAN J, et al Mergingthermal and Visual Images by a Contrast Pyramid[J].Optical En-gineering,1989(7):789-792.
[34]DUFOURNAUD Y, SCHMID C, HORAUD R. Image Matching with Scale Adjustment[J]. Computer Vision and Image Understanding,2004,93(2):175-194.
[35]陈鹰,林怡.基于提升小波的影像变换与匹配[J].测绘学报,2006,35(1)：19-23.
[36]ATALLAH M J. Faster Image Template Matching in the Sum of the Absolute Value of Differences Measure[J].IEEE Transactions on Image Processing,2001,10(4):659-663.
[37]SHAMIR L. A Proposed Stereo Matching Algorithm for Noisy Sets of Color Images[J]. Computers&Geosciences,2007,33(8):1052-1063.
[38]江万寿,郑顺义,张祖勋,等.航空影像特征匹配研究[J].武汉大学学报：信息科学版,2003,28(5)：510-513.
[39]ANSARI M E, MASMOUDI L, BENSRHAIR A. A New Regions Matching for Color Stereo Images [J]. Pattern Recognition Letters,2007,28(13):1679-1687.
[40]ZULIANI M, BERTELLI L, KENNEY C S,et al. Drums,Curve Descriptors and Affine Invariant Region Matching [J]. Image and Vision Computing,2008,26(3):347-360.
[41]纪华,吴元昊,孙宏海,等.结合全局信息的SIFT特征匹配算法[J].光学精密工程,2009,17(2)：439-444.
[42]David G.Lowe.Object recognition from local scale-invariant features[J].International Conference on Computer Vision,1999,3(1):1150-1157.
[43]David G.Lowe.Distinctive Image Features from Scale-Invarient Keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[44]柯涛.旋转多基线数字近景摄影测量[D].武汉：武汉大学博士学位论文,2008.
[45]Mikolajczyk K, Schimid C.A Performance Evaluation of Local Descriptors [J]. IEEE Trans Pattern Analysis and Machine Intelligence,2005,27(10):1615-1630.
[46]Friedman.J.H, Bentley.J.L, and Finkel.R.A.An algorithm for finding best matches in logarithmic expected time[J].ACM Transations on Mathematical software,1977,3(3):209-226.
[47]谢萍,邹峥嵘,肖奇.基于Harris角点和SIFT特征的近景影像匹配[J].测绘科学,2011.
[48]M. A. Fischler and R. C. Bolles."Random Sample Consensus:A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography". Comm. Of the ACM24(6):381-395,1981.
[49]GBT_23236-2009,数字航空摄影测量_空中三角测量规范[S].国家测绘局,2009.
[50]CHZ3003-2010,低空数字航空摄影测量内业规范[S].国家测绘局,2010.
[2]孙杰,林宗坚,崔红霞.无人机低空遥感监测系统[J].遥感信息,2003,(1)：4950,27.
[3]Sang-Jong Lee, Seong-pil Kim, Tae-Sik Kim, etal. Development of Autonomous Flight Control System for50m Unmanned Airship [J]. IS SNIP,2004,457-462.
[4]Filoktimon Repoulias, Evangelos Papadopoulos. Dynamically Feasible Trajectory and Open-Loop Control Design for Unmanned Airships[C].2007Mediterranean Conference on Control and Automation,2007.
[5]Jose Raul Azinheira, Ely Carneiro de Paiva, Josue Jr.G.Ramos, et al. Mission Path Following for an Autonomous Unmanned Airship [C]. International Conference on Robotics&Automation,2000.
[6]D.Gray, M.Nagatsuka, R.Miura. Autopsy of Unmanned Airship Television Broadcast [J]. Yokosuka Radio Communications Research Centre,2006.
[7]Herwitz R.S., Lee F.J., Arvesen J.C. Precision Agriculture as a Commercial Application for Solar-Powered Unmanned Aerial Vehicles [C].1st American Institue of Aeronautics and Astronautics UAV Conference,2002.
[8]Jinjun Rao, Zhenbang Gong, Jun Luo, et al. Unmanned Airships for Emergency Management[C]. International Workshop on Safety, Security and Rescue Robotics,2005.
[9]Dugan J. and Piotrowski C. Develpomental system for maritime rapid environment assessment using UAVs[C].5st ONERA_DLR aerospace symposium,2003.
[10]Robert M.N. Unmanned air vehicles-coming of age at last[J]. Air&space Europe,2000,2(1):23-25.
[11]Toth C.K, G-Brzezinska D., Merry C. Supporting traffic flow management with high-definition imagery[C]. ISPRS workshop on high resolution mapping from space2003. Hannover,2003.
[12]戴斌.航空摄影测量在土地利用现状更新调查中的应用[J].铁道勘察,2005,(6)：24-26.
[13]申海建,郭荣中,黄小波等.微型无人机(MUAV)航空摄影测量技术在土地整理项目规划设计中的应用[C].2007年中国土地学会年会,2007.
[14]张雪萍,刘英.无人机在大比例尺DOM生产中的应用[J].测绘标准化,2011,27(4)：25-27.
[15]李雷,陈桂琴,田园.无人机的发展历程与展望[J].科技论坛.
[16]http://chzt.sbsm.gov.cn/article/qt/xjsxcpcxyy/ckxt/jsjs/.新技术新产品创新应用.
[17]张静.“光照度平方反比定律”的实验设计[J].物理教学探讨,2009,(19)
[18]CJJ899,城市测量规范[S].
[19]林君建,苍桂华.摄影测量学[M].北京：国防工业出版社,2005.
[20]王青山.简述无人机在遥感技术中的应用[J],测绘与空间地理信息,2010,(3),101.
[21]马瑞升,孙涵,林宗桂,等.微型无人机遥感影像的纠偏与定位[J].南京气象学院学报,2005,28(5)：632-639.
[22]王俊.无人机航空摄影的空三评价分析[J].甘肃科技,2011,13(27)：40-43.
[23]宫阿都,何孝莹,雷添杰,李京.无控制点数据的无人机影像快速处理[J].地球信息科学学报,2010,2(12)：254-259.
[24]程效军,胡敏捷.数码相机畸变差的检测[J].测绘学报,2002,(31)：113-117.
[25]张祖勋,张剑清.数字摄影测量学.武汉：武汉测绘科技大学出版社,1997.216-232.
[26]宣家斌.航摄仪内方位元素和物镜畸变差的测定
[27]邹勇平,徐水平,李月锋,林沂才.实验室检定航摄相机内方位元素和畸变差计算方法的比较[J].测绘技术装备,2006,1：37-38.
[28]崔红霞,孙杰,林宗坚,储美华.非量测数码相机的畸变差检测研究[J].测绘科学,2005,1(30)：105-107.
[29]金仲华,崔红霞.数码相机径向畸变差检测及改正[J].科技广场,2008.
[30]冯文灏,商浩亮.成像系统通用数字畸变模型的建立原理与应用.数字近景摄影测量会议,武汉,2004.
[31]C.J.Harris, M.StePhens.A combined Corner and Edge Detector[C]. Proceedings of the Fourth Alvey Vision Conference,1988:147-151.
[32]王玉宏,陈永立.低空遥感影像匹配与拼接技术研究.测绘信息与工程,2009,34(2).
[33]TOET A, RUYVEN L, VAN J, et al Mergingthermal and Visual Images by a Contrast Pyramid[J].Optical En-gineering,1989(7):789-792.
[34]DUFOURNAUD Y, SCHMID C, HORAUD R. Image Matching with Scale Adjustment[J]. Computer Vision and Image Understanding,2004,93(2):175-194.
[35]陈鹰,林怡.基于提升小波的影像变换与匹配[J].测绘学报,2006,35(1)：19-23.
[36]ATALLAH M J. Faster Image Template Matching in the Sum of the Absolute Value of Differences Measure[J].IEEE Transactions on Image Processing,2001,10(4):659-663.
[37]SHAMIR L. A Proposed Stereo Matching Algorithm for Noisy Sets of Color Images[J]. Computers&Geosciences,2007,33(8):1052-1063.
[38]江万寿,郑顺义,张祖勋,等.航空影像特征匹配研究[J].武汉大学学报：信息科学版,2003,28(5)：510-513.
[39]ANSARI M E, MASMOUDI L, BENSRHAIR A. A New Regions Matching for Color Stereo Images [J]. Pattern Recognition Letters,2007,28(13):1679-1687.
[40]ZULIANI M, BERTELLI L, KENNEY C S,et al. Drums,Curve Descriptors and Affine Invariant Region Matching [J]. Image and Vision Computing,2008,26(3):347-360.
[41]纪华,吴元昊,孙宏海,等.结合全局信息的SIFT特征匹配算法[J].光学精密工程,2009,17(2)：439-444.
[42]David G.Lowe.Object recognition from local scale-invariant features[J].International Conference on Computer Vision,1999,3(1):1150-1157.
[43]David G.Lowe.Distinctive Image Features from Scale-Invarient Keypoints[J].International Journal of Computer Vision,2004,60(2):91-110.
[44]柯涛.旋转多基线数字近景摄影测量[D].武汉：武汉大学博士学位论文,2008.
[45]Mikolajczyk K, Schimid C.A Performance Evaluation of Local Descriptors [J]. IEEE Trans Pattern Analysis and Machine Intelligence,2005,27(10):1615-1630.
[46]Friedman.J.H, Bentley.J.L, and Finkel.R.A.An algorithm for finding best matches in logarithmic expected time[J].ACM Transations on Mathematical software,1977,3(3):209-226.
[47]谢萍,邹峥嵘,肖奇.基于Harris角点和SIFT特征的近景影像匹配[J].测绘科学,2011.
[48]M. A. Fischler and R. C. Bolles."Random Sample Consensus:A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography". Comm. Of the ACM24(6):381-395,1981.
[49]GBT_23236-2009,数字航空摄影测量_空中三角测量规范[S].国家测绘局,2009.
[50]CHZ3003-2010,低空数字航空摄影测量内业规范[S].国家测绘局,2010.