空中三角测量中航线间公共点的自动提取技术研究
|
摘要
在航空摄影过程中,由于摄影时间的变化、太阳高度角的变化,使的地面物体和有一定高度的地面植被在不同航线上的位置发生变化,而在一般的纹理分析,特征点提取时,这些点很容易被提取出来,这就造成了点位的不精确,降低了空中三角测量的精度。实践证明,要克服摄影时间、太阳高度角、摄影位置等变化因素的影响,必须提取一些特殊的特征点及相应的匹配方法,这是解决全自动空中三角测量的关键。本文研究了特征点提取的相关算法,目的是解决空中三角测量中航线公共点的自动提取与匹配问题。具体研究内容包括以下四个方面。
1.图像预处理
介绍了一种Wallis变换的图像增强方法,针对其在影像上局部分块处理造成的块效应影响,提出一种基于双线性插值的Wallis变换方法,有效克服了块效应现象。通过基于LoG算子的blob特征检测与初始匹配的实验,证明该方法有利于特征提取和伪特征剔除。
2.现有特征点提取方法和匹配方法
介绍了摄影测量领域中应用比较广泛的几种特征点提取算子:Moravec算子、Forstner算子、Harris算子以及SUSAN算子,通过大量的实验从多方面综合比较了各算子的性能,指出各自的优缺点及适应环境。还介绍了不同的匹配方法,从传统的归一化相关系数匹配,到高精度的单点最小二乘匹配,另外设计了SIFT特征匹配方法及其在航空摄影测量中的应用,实验证明此方法只能用于粗略定向。
3.本文航线连接点提取方法及匹配实验
重点介绍了blob特征检测方法,并在此方法的基础上根据空中三角测量中航线公共点的特殊性,经过相关理论知识和参数适应性,提出了一种适合本文的特殊点的提取方法,通过对相邻像对进行归一化的相关系数匹配和单点最小二乘匹配,以及相对定向的误匹配剔除,最后进行相对定向,证明了该方法能克服了传统方法对摄影时间、太阳高度角等外界环境的干扰,最后对二个测区的航线间公共点的自动提取与匹配进行评估实验。
In the aerial photography process, because of the changes of photograph time、solar elevation angle、some ground objects and ground vegetation with certain height have different locations on different routes, and for the general texture analysis,in the process of feature extraction,These points are easily extracted, which resulted in imprecise points, reducing the accuracy of aerial triangulation.Practice proved to overcome the impact caused by the photography time, sun elevation angle, photography locations,some special feature points must be extracted and the corresponding feature matching method must be applied, which is the key for automatic aerial triangulation. In this paper, the relevant feature extraction algorithms are studied with the aim to realize the extraction and matching of the tie point in aerial triangulation .The main contents and conclusions of this paper are as follows:
1.the image preprocessing
Wallis transformation is widely employed for image enhancement in the field of photogrammety and remote sensing. However, the original wallis transformation has the blocking effect phenomenon, which will yield many false features in image matching. In this paper, an improved wallis transformation is proposed, in which the bilinear interpolation is adopted to avoid the blocking effect. The experiment about blob extraction and image matching, using the LoG operator, demonstrates that our method is useful to increase the number of features extracted and decrease the number of false features.
2.the traditional feature points extraction and match
Several feature extraction operators are introduced which have a wide application in photogrammetry : Moravec operator, Forstner operator, Harris operator and the SUSAN operator, the paper gives a performance through a large number of experiments,including the advantages and disadvantages and the adapted conditions.
The SIFT feature matching method and the application in photogrammetry is studied, SIFT features are extracted in images reduced in size,the mismatch points are removed using the geometric constraint, and relocate the features to the original image with the pyramid-related method , the experiment proved that the method can only be used for rough location for aerial image and not fit for the aerial triangulation which need high precision.
3.the extraction and matching of the tie point in aerial triangulation
The method of extraction of the tie point is proposed based on the theory of LoG,The correlation matching, single point least squares matching and relative orientation are explained;Finally, the experiments of relative orientation proved that the Blob features automatically extracted meet the requirement of relative orientation.Also,the paper give a access of the method using two arrears,and the match number and precision is good,the method help to resolve the problem.
1.图像预处理
介绍了一种Wallis变换的图像增强方法,针对其在影像上局部分块处理造成的块效应影响,提出一种基于双线性插值的Wallis变换方法,有效克服了块效应现象。通过基于LoG算子的blob特征检测与初始匹配的实验,证明该方法有利于特征提取和伪特征剔除。
2.现有特征点提取方法和匹配方法
介绍了摄影测量领域中应用比较广泛的几种特征点提取算子:Moravec算子、Forstner算子、Harris算子以及SUSAN算子,通过大量的实验从多方面综合比较了各算子的性能,指出各自的优缺点及适应环境。还介绍了不同的匹配方法,从传统的归一化相关系数匹配,到高精度的单点最小二乘匹配,另外设计了SIFT特征匹配方法及其在航空摄影测量中的应用,实验证明此方法只能用于粗略定向。
3.本文航线连接点提取方法及匹配实验
重点介绍了blob特征检测方法,并在此方法的基础上根据空中三角测量中航线公共点的特殊性,经过相关理论知识和参数适应性,提出了一种适合本文的特殊点的提取方法,通过对相邻像对进行归一化的相关系数匹配和单点最小二乘匹配,以及相对定向的误匹配剔除,最后进行相对定向,证明了该方法能克服了传统方法对摄影时间、太阳高度角等外界环境的干扰,最后对二个测区的航线间公共点的自动提取与匹配进行评估实验。
In the aerial photography process, because of the changes of photograph time、solar elevation angle、some ground objects and ground vegetation with certain height have different locations on different routes, and for the general texture analysis,in the process of feature extraction,These points are easily extracted, which resulted in imprecise points, reducing the accuracy of aerial triangulation.Practice proved to overcome the impact caused by the photography time, sun elevation angle, photography locations,some special feature points must be extracted and the corresponding feature matching method must be applied, which is the key for automatic aerial triangulation. In this paper, the relevant feature extraction algorithms are studied with the aim to realize the extraction and matching of the tie point in aerial triangulation .The main contents and conclusions of this paper are as follows:
1.the image preprocessing
Wallis transformation is widely employed for image enhancement in the field of photogrammety and remote sensing. However, the original wallis transformation has the blocking effect phenomenon, which will yield many false features in image matching. In this paper, an improved wallis transformation is proposed, in which the bilinear interpolation is adopted to avoid the blocking effect. The experiment about blob extraction and image matching, using the LoG operator, demonstrates that our method is useful to increase the number of features extracted and decrease the number of false features.
2.the traditional feature points extraction and match
Several feature extraction operators are introduced which have a wide application in photogrammetry : Moravec operator, Forstner operator, Harris operator and the SUSAN operator, the paper gives a performance through a large number of experiments,including the advantages and disadvantages and the adapted conditions.
The SIFT feature matching method and the application in photogrammetry is studied, SIFT features are extracted in images reduced in size,the mismatch points are removed using the geometric constraint, and relocate the features to the original image with the pyramid-related method , the experiment proved that the method can only be used for rough location for aerial image and not fit for the aerial triangulation which need high precision.
3.the extraction and matching of the tie point in aerial triangulation
The method of extraction of the tie point is proposed based on the theory of LoG,The correlation matching, single point least squares matching and relative orientation are explained;Finally, the experiments of relative orientation proved that the Blob features automatically extracted meet the requirement of relative orientation.Also,the paper give a access of the method using two arrears,and the match number and precision is good,the method help to resolve the problem.
引文
[1] H. P. Moravec. Towards Automatic Visual Obstacle Avoidance. Proc. 5th International Joint Conference on Artificial Intelligence, pp. 584, 1977.
[2] H. P. Moravec. Visual Mapping by a Robot Rover. International Joint Conference on Artificial Intelligence, pp. 598-600, 1979.
[3] Beaudet,P,1978:Rotationally invariant image operators.Proc.4th Internaltional Joint Conference on Pattern Recognition,pp.579-583.
[4] Mikolajczyk K,Schmid C.Sacle&affine invariant interest point detectors[J].International Journal of Computer Vision 2004.60(1):63-86.
[5] Kitchen,L.,Rosenfeld,A.,1982.Gray-level corner detection.Pattern Recognition Letters 195-102.
[6] W. F?rstner and E. Gülch, A fast operator for detection and precise location of distinct points, corners and centers of circular features. In: ISPRS Intercommision Workshop (June 1987), pp. 149–155.
[7] Harris C G, Stephens M J. A Combined Corner and Edge Detector [C]. Manchester: Manchester Proceedings Fourth Alvey Vision Conference,1988.
[8] Wang,H.,Brady,M.,1995.Real-time corner detection algorithm for motion estimation.Image and Vision Computing 13(9).
[9] Smith,S.M.and Brady,J.M.,1997:SUSAN–a new approach to low level image processing,Int.Journal of Computer Vision, Vol.23,No.1,pp.45-78.
[10] M. Trajkovic and M. Hedley. Fast Corner Detection. Image and Vision Computing, Vol. 16(2), pp. 75-87, 1998.
[11] F. Mokhtarian and R. Suomela, Robust Image Corner Detection Trough Curvature Scale Space, TPAMI, Vol. 20, No. 12, 1998.
[12] Z.Zheng, H.Wang and EKTeoh. Analysis of Gray Level Corner Detection. Pattern Recognition Letters, Vol. 20, pp. 149-162, 1999.
[13] T.Lindeberg.Scale-Spcce Theory in Computer Vision.Kluwer Publishers,1994.
[14] Linderberg T.Feature detection with automatic scale selection[J].InternationalJournal of Computer Vision 1998.30(2):79-116
[15] Y.Dufournaud,C.Schmid,and R.Horaud.Matching images with different resolutions. In Proceedings of the Conference on Computer Vision and Pattern Recognition,Hilton Head Island,South Carolina,USA,pages 612–618,2000.
[16] G.Lowe D.Object Recognition from Local Scale-Invariant Features[C].In:Proceeding of the International Conference on Computer Vision(ICCV’1999).1999.
[17] D.G.Lowe.Distinctive image features from scale-invariant keypoints.International Journal of Computer Vision,60(2):91-110,2004.
[18] K.Mikolajczyk and C.Schmid.Indexing based on scale invariant interest points.InProceedings of the 8th International Conference on Computer Vision,Vancouver,Canada,pages 525–531,2001.
[19] Mikolajczyk K,Schmid C.Sacle&affine invariant interest point detectors[J].International Journal of Computer Vision 2004.60(1):63-86.
[20] T.Tuytelaars and L.Van Gool.Wide baseline stereo matching based on local affnely invariant regions.In The Eleventh British Machine Vision Conference,University of Bristol,UK,pages 412–425,2000.
[21] Matas J,Chum O,Urban M,Pajdla T.Robust wide-baseline stereo from maximally stable extremal regions[C].In:Proceedings of the British Machine Vision Conference.2002.
[22] Kadir T,Brady M,Zisserman A.An affine invariant method for selecting salient regions in images[C].In:the 8th European Conference on Computer Vision.2004.Tahlequah Republic.
[23] A.Johnson and M.Hebert,"Object Recognition by Matching Oriented Points,"Proc.Conf.ComputerVision and Pattern Recognition,pp.684-689,1997.
[24] S.Lazebnik,C.Schmid,and J.Ponce,"Sparse Texture Representation Using Affine-Invariant Neighborhoods,"Proc.Conf.Computer Vision and Pattern Recognition,pp.319-324,2003.
[25] R.Zabih and J.Woodfill,"Non-Parametric Local Transforms for Computing Visual Correspondance,"Proc.Third European Conf.Computer Vision,pp.151-158,1994.
[26] Y.Ke and R.Sukthankar.PCA-SIFT:A More Distinctive Representation for Local Image descriptors.Proe.Conf.Computer Vision and Patten Recognition.PP.511-517, 2004.
[27] Mikolajczyk K,Schmid C.A performance evaluation of local descriptors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence 2005.27(10):1615-1630.
[28] M.Brown,R,Szeliski and S.Winder. Multi-image matching using multi-scale oriented Patches.SanDiego,Calif,USA:In Proeeedings of the IEEE Computer Soeiety Conference on Computer Vision and Pattern Recognition,2005,l:510-517.
[29]孙坚伟,王汝笠.改进的MOPs图像匹配算法.科学技术与工程.2006,116(2l).
[30] S.Lazebnik,C.Schmid,and J.Ponee.A sparse texture representation using local affine regions.In CVPR,Pages309-2003.
[31] H.Bay,T.Tuytelaars,and L.vanGool.SURF:Speeded up robust features.In ECCV,2006 to appear.
[32] J.Koenderink,A.van Doorn. Representation of local geometry in the visual system.Biological Cybemeties,55:367es375,1987.
[33] L.Florack,B.ter Haar Romeny, J.Koenderink,and M.Viergever.General intensity transformations and second order invariants.In Proceedings of the 7th Scandinavian Conference on Image Analysis,Aalborg,Denmark,pages 338-345,1991.
[34] W.Freeman , E.Adelson.The design and useof steerable filters.IEEE Transactions on Partern Analysis and Machine Intelligence,13(9):891-906,1991.
[35] A.Baumberg.Reliable feature matching across widely separated views.In Proceedings of The Conference on Computer Vision and Pattern Recognition, HiltonHeadIsland,South Carolina,USA,Pages774-781,2000.
[36] F.Schaffalitzky and A.Zisserman.Multi-view matching for unordered image sets.In ECCV,pp.414-431,2002.
[37]张力,张祖勋,张剑清.Wallis滤波在影像匹配中的应用[J].武汉测绘科技大学学报,1999,24(1):24-27.
[38]王智均,李德仁,李清泉.Wallis变换在小波影像融合中的应用[J].武汉测绘科技大学学报,2000,25(4):338-342.
[39]李德仁,王密,潘俊.光学遥感影像的自动匀光处理及应用[J].武汉大学·信息科学版,2006,31(9):753-756.
[40]王密,潘俊.一种数字航空影像的匀光方法[J].中国图像图报,2004,9(6): 744-748.
[41]冈萨雷斯.数字图像处理[M].北京:电子工业出版社, 2008: 105-106.
[42] Lindeberg,T.1994.Scale-space theory:A basic tool for analysing structures at different scales.Journal of Applied statistics,21(2):224-270
[43] Fishchler M A:Random Sample Consensus:a Paradigm for model fitting with application to image analysis automated cartography[J].Communication Association Machine,1981,24(6):381-395
[2] H. P. Moravec. Visual Mapping by a Robot Rover. International Joint Conference on Artificial Intelligence, pp. 598-600, 1979.
[3] Beaudet,P,1978:Rotationally invariant image operators.Proc.4th Internaltional Joint Conference on Pattern Recognition,pp.579-583.
[4] Mikolajczyk K,Schmid C.Sacle&affine invariant interest point detectors[J].International Journal of Computer Vision 2004.60(1):63-86.
[5] Kitchen,L.,Rosenfeld,A.,1982.Gray-level corner detection.Pattern Recognition Letters 195-102.
[6] W. F?rstner and E. Gülch, A fast operator for detection and precise location of distinct points, corners and centers of circular features. In: ISPRS Intercommision Workshop (June 1987), pp. 149–155.
[7] Harris C G, Stephens M J. A Combined Corner and Edge Detector [C]. Manchester: Manchester Proceedings Fourth Alvey Vision Conference,1988.
[8] Wang,H.,Brady,M.,1995.Real-time corner detection algorithm for motion estimation.Image and Vision Computing 13(9).
[9] Smith,S.M.and Brady,J.M.,1997:SUSAN–a new approach to low level image processing,Int.Journal of Computer Vision, Vol.23,No.1,pp.45-78.
[10] M. Trajkovic and M. Hedley. Fast Corner Detection. Image and Vision Computing, Vol. 16(2), pp. 75-87, 1998.
[11] F. Mokhtarian and R. Suomela, Robust Image Corner Detection Trough Curvature Scale Space, TPAMI, Vol. 20, No. 12, 1998.
[12] Z.Zheng, H.Wang and EKTeoh. Analysis of Gray Level Corner Detection. Pattern Recognition Letters, Vol. 20, pp. 149-162, 1999.
[13] T.Lindeberg.Scale-Spcce Theory in Computer Vision.Kluwer Publishers,1994.
[14] Linderberg T.Feature detection with automatic scale selection[J].InternationalJournal of Computer Vision 1998.30(2):79-116
[15] Y.Dufournaud,C.Schmid,and R.Horaud.Matching images with different resolutions. In Proceedings of the Conference on Computer Vision and Pattern Recognition,Hilton Head Island,South Carolina,USA,pages 612–618,2000.
[16] G.Lowe D.Object Recognition from Local Scale-Invariant Features[C].In:Proceeding of the International Conference on Computer Vision(ICCV’1999).1999.
[17] D.G.Lowe.Distinctive image features from scale-invariant keypoints.International Journal of Computer Vision,60(2):91-110,2004.
[18] K.Mikolajczyk and C.Schmid.Indexing based on scale invariant interest points.InProceedings of the 8th International Conference on Computer Vision,Vancouver,Canada,pages 525–531,2001.
[19] Mikolajczyk K,Schmid C.Sacle&affine invariant interest point detectors[J].International Journal of Computer Vision 2004.60(1):63-86.
[20] T.Tuytelaars and L.Van Gool.Wide baseline stereo matching based on local affnely invariant regions.In The Eleventh British Machine Vision Conference,University of Bristol,UK,pages 412–425,2000.
[21] Matas J,Chum O,Urban M,Pajdla T.Robust wide-baseline stereo from maximally stable extremal regions[C].In:Proceedings of the British Machine Vision Conference.2002.
[22] Kadir T,Brady M,Zisserman A.An affine invariant method for selecting salient regions in images[C].In:the 8th European Conference on Computer Vision.2004.Tahlequah Republic.
[23] A.Johnson and M.Hebert,"Object Recognition by Matching Oriented Points,"Proc.Conf.ComputerVision and Pattern Recognition,pp.684-689,1997.
[24] S.Lazebnik,C.Schmid,and J.Ponce,"Sparse Texture Representation Using Affine-Invariant Neighborhoods,"Proc.Conf.Computer Vision and Pattern Recognition,pp.319-324,2003.
[25] R.Zabih and J.Woodfill,"Non-Parametric Local Transforms for Computing Visual Correspondance,"Proc.Third European Conf.Computer Vision,pp.151-158,1994.
[26] Y.Ke and R.Sukthankar.PCA-SIFT:A More Distinctive Representation for Local Image descriptors.Proe.Conf.Computer Vision and Patten Recognition.PP.511-517, 2004.
[27] Mikolajczyk K,Schmid C.A performance evaluation of local descriptors[J].IEEE Transactions on Pattern Analysis and Machine Intelligence 2005.27(10):1615-1630.
[28] M.Brown,R,Szeliski and S.Winder. Multi-image matching using multi-scale oriented Patches.SanDiego,Calif,USA:In Proeeedings of the IEEE Computer Soeiety Conference on Computer Vision and Pattern Recognition,2005,l:510-517.
[29]孙坚伟,王汝笠.改进的MOPs图像匹配算法.科学技术与工程.2006,116(2l).
[30] S.Lazebnik,C.Schmid,and J.Ponee.A sparse texture representation using local affine regions.In CVPR,Pages309-2003.
[31] H.Bay,T.Tuytelaars,and L.vanGool.SURF:Speeded up robust features.In ECCV,2006 to appear.
[32] J.Koenderink,A.van Doorn. Representation of local geometry in the visual system.Biological Cybemeties,55:367es375,1987.
[33] L.Florack,B.ter Haar Romeny, J.Koenderink,and M.Viergever.General intensity transformations and second order invariants.In Proceedings of the 7th Scandinavian Conference on Image Analysis,Aalborg,Denmark,pages 338-345,1991.
[34] W.Freeman , E.Adelson.The design and useof steerable filters.IEEE Transactions on Partern Analysis and Machine Intelligence,13(9):891-906,1991.
[35] A.Baumberg.Reliable feature matching across widely separated views.In Proceedings of The Conference on Computer Vision and Pattern Recognition, HiltonHeadIsland,South Carolina,USA,Pages774-781,2000.
[36] F.Schaffalitzky and A.Zisserman.Multi-view matching for unordered image sets.In ECCV,pp.414-431,2002.
[37]张力,张祖勋,张剑清.Wallis滤波在影像匹配中的应用[J].武汉测绘科技大学学报,1999,24(1):24-27.
[38]王智均,李德仁,李清泉.Wallis变换在小波影像融合中的应用[J].武汉测绘科技大学学报,2000,25(4):338-342.
[39]李德仁,王密,潘俊.光学遥感影像的自动匀光处理及应用[J].武汉大学·信息科学版,2006,31(9):753-756.
[40]王密,潘俊.一种数字航空影像的匀光方法[J].中国图像图报,2004,9(6): 744-748.
[41]冈萨雷斯.数字图像处理[M].北京:电子工业出版社, 2008: 105-106.
[42] Lindeberg,T.1994.Scale-space theory:A basic tool for analysing structures at different scales.Journal of Applied statistics,21(2):224-270
[43] Fishchler M A:Random Sample Consensus:a Paradigm for model fitting with application to image analysis automated cartography[J].Communication Association Machine,1981,24(6):381-395