犯罪案件现场虚拟重建技术研究
|
摘要
犯罪现场勘查,是指侦查人员依据法律规定,对与犯罪有关的时间、地点和人、事、物进行的现场调查、勘验检查、现场记录等侦查活动,对于查明案件性质、收集犯罪证据、分析案犯特点、确定侦查方向具有十分重要的作用,是侦查活动的起点,也是公安刑侦工作的一个重要组成部分。传统的现场勘查工作做法是通过现场笔录、现场图、现场照片、现场录像、各类鉴定结论以及现场勘验人员的口头描述等方法来综合反映现场信息。这些方法存在信息反映的主观性、局限性,能够获取到的信息量多寡也因人而异。
当前,计算机科学与工程技术飞速发展,已经广泛应用于各行各业。如何将其与现场勘查技术有机结合起来,提高现场勘查工作的信息化、科学化水平,成为侦查工作中充分发挥现场及物证作用的当务之急。本文紧紧围绕公安部门现场勘查工作的特点和需求,结合虚拟现实、图像处理等技术,提出了一种基于IBR(Image Based Rendering,基于图像渲染)技术的犯罪案件现场虚拟重建方法,并重点对现场全景图像生成过程中的若干关键技术进行了研究。
论文首先对现场全景图像生成过程建立了总体模型,将整个问题分为现场鱼眼图像采集、鱼眼图像校正恢复、现场图像预处理、现场图像匹配和图像平滑五个模块,使课题的研究过程变得清晰明了。在此基础上,对每个模块分别搭建了研究框架,使得研究过程的每一个环节,都紧扣现场环境的特点,在案件现场虚拟重建这一总体模型下来进行。
深入研究了鱼眼镜头的构造和成像原理,从现场环境和鱼眼镜头的特点以及问题研究的需求出发,对图像采集过程中的技术细节进行了规范。并使用基于球面透视投影约束的方法,对采集的现场鱼眼图像进行了校正,较好地恢复了鱼眼图像的桶形畸变。
系统分析了图像预处理技术的一些基本原理和方法,根据案件现场的特点和现场重建的需求,综合运用了直方图规定化和中值滤波两种方法对现场图像进行了匹配前的预处理,改善了图像的视觉效果、减少了噪声影响。
针对犯罪案件现场重建过程中的图像匹配环节,提出了一种精确的匹配算法。该算法提取的特征点分布均匀、稳定可靠,虽然匹配的速度不是很快,但是匹配精确、稳定,非常适合对匹配实时性要求不高但准确性要求高的案件现场图像进行匹配。
在图像匹配完成后,采用一种渐入渐出的算法对图像进行平滑处理,达到了较好的效果。
The survey of the criminal scenes means the personnel makes spot survey, investigation and record on the time, place, people, accidents and contents related to the crime. It is significant to find out the character of the case, collect the evidence of the criminal, analyze the characteristic of the criminal and confirm the direction of the detection, which is the starting point of detection and also an important part of police criminal detection. It is the traditional methods to reflect information of the scene through deposition, field images, field photos, field videos, different survey conclusions and oral descriptions of detection personnel. These methods contain subjectivity and localization of the information and the amount of information acquired differs by different person.
Nowadays, computer science and engineering technology develops rapidly and has been applied in many professions. It is an urgent affair of how to combine it to scene detection, which can improve the information and science levels of scene detection. This paper encircled the characteristics and requirements of scene detection of police department, presented a virtual rebuilding method based on IBR (Image Based Rendering), which associated with virtual reality and image processing. And made emphases research on some key techniques of panorama image creation.
Firstly, it made a universe model of the creating process of scene panorama image and divided the whole issue to five modules, which were scene fisheye image acquisition, fisheye lens distortion correction, pre-processing of the scene image, scene image matching after pre-processing and scene image smoothing. Secondly, it built research sub-models separately of each module, which made each segment of research process coincide the characteristic of scene circumstance and carry through in the frame of criminal scene virtual rebuilding.
It had made deep research of the constitution and imaging principle of fisheye lens and made criterion of detail technique during image acquisition on the character of scene circumstance and fisheye lens and the demand of issues investigation.
It summarized some methods of image pre-processing technique systematically. Histogram and mid-value filtering were applied to pre-process the scene image, which developed the sight effect and reduced the influence of noises.
A precise matching algorithm was put forward against the image matching segment of criminal scene rebuilding. The characteristic points extracted by the algorithm distributed equally, stably and credibly. Although the matching speed did not fast, it matched precisely and stably. It was very suitable on the criminal scene image matching where need general real time quality, but high accuracy. It was also the main innovate point in this paper.
After image matching completed, we adopt a type of fade in fade out algorithm on the images to smooth and achieved preferable results.
当前,计算机科学与工程技术飞速发展,已经广泛应用于各行各业。如何将其与现场勘查技术有机结合起来,提高现场勘查工作的信息化、科学化水平,成为侦查工作中充分发挥现场及物证作用的当务之急。本文紧紧围绕公安部门现场勘查工作的特点和需求,结合虚拟现实、图像处理等技术,提出了一种基于IBR(Image Based Rendering,基于图像渲染)技术的犯罪案件现场虚拟重建方法,并重点对现场全景图像生成过程中的若干关键技术进行了研究。
论文首先对现场全景图像生成过程建立了总体模型,将整个问题分为现场鱼眼图像采集、鱼眼图像校正恢复、现场图像预处理、现场图像匹配和图像平滑五个模块,使课题的研究过程变得清晰明了。在此基础上,对每个模块分别搭建了研究框架,使得研究过程的每一个环节,都紧扣现场环境的特点,在案件现场虚拟重建这一总体模型下来进行。
深入研究了鱼眼镜头的构造和成像原理,从现场环境和鱼眼镜头的特点以及问题研究的需求出发,对图像采集过程中的技术细节进行了规范。并使用基于球面透视投影约束的方法,对采集的现场鱼眼图像进行了校正,较好地恢复了鱼眼图像的桶形畸变。
系统分析了图像预处理技术的一些基本原理和方法,根据案件现场的特点和现场重建的需求,综合运用了直方图规定化和中值滤波两种方法对现场图像进行了匹配前的预处理,改善了图像的视觉效果、减少了噪声影响。
针对犯罪案件现场重建过程中的图像匹配环节,提出了一种精确的匹配算法。该算法提取的特征点分布均匀、稳定可靠,虽然匹配的速度不是很快,但是匹配精确、稳定,非常适合对匹配实时性要求不高但准确性要求高的案件现场图像进行匹配。
在图像匹配完成后,采用一种渐入渐出的算法对图像进行平滑处理,达到了较好的效果。
The survey of the criminal scenes means the personnel makes spot survey, investigation and record on the time, place, people, accidents and contents related to the crime. It is significant to find out the character of the case, collect the evidence of the criminal, analyze the characteristic of the criminal and confirm the direction of the detection, which is the starting point of detection and also an important part of police criminal detection. It is the traditional methods to reflect information of the scene through deposition, field images, field photos, field videos, different survey conclusions and oral descriptions of detection personnel. These methods contain subjectivity and localization of the information and the amount of information acquired differs by different person.
Nowadays, computer science and engineering technology develops rapidly and has been applied in many professions. It is an urgent affair of how to combine it to scene detection, which can improve the information and science levels of scene detection. This paper encircled the characteristics and requirements of scene detection of police department, presented a virtual rebuilding method based on IBR (Image Based Rendering), which associated with virtual reality and image processing. And made emphases research on some key techniques of panorama image creation.
Firstly, it made a universe model of the creating process of scene panorama image and divided the whole issue to five modules, which were scene fisheye image acquisition, fisheye lens distortion correction, pre-processing of the scene image, scene image matching after pre-processing and scene image smoothing. Secondly, it built research sub-models separately of each module, which made each segment of research process coincide the characteristic of scene circumstance and carry through in the frame of criminal scene virtual rebuilding.
It had made deep research of the constitution and imaging principle of fisheye lens and made criterion of detail technique during image acquisition on the character of scene circumstance and fisheye lens and the demand of issues investigation.
It summarized some methods of image pre-processing technique systematically. Histogram and mid-value filtering were applied to pre-process the scene image, which developed the sight effect and reduced the influence of noises.
A precise matching algorithm was put forward against the image matching segment of criminal scene rebuilding. The characteristic points extracted by the algorithm distributed equally, stably and credibly. Although the matching speed did not fast, it matched precisely and stably. It was very suitable on the criminal scene image matching where need general real time quality, but high accuracy. It was also the main innovate point in this paper.
After image matching completed, we adopt a type of fade in fade out algorithm on the images to smooth and achieved preferable results.
引文
[1]胡晓峰,李国辉.多媒体系统[M].北京:人民邮电出版社,1997.
[2]张茂军.虚拟现实系统[M].北京:科学出版社,2001.
[3]杨东录.犯罪现场勘查实用技术[M].西安:陕西人民出版社,2003.
[4]S.E.Chen, L.Williams. View Interpolation for Image Synthesis[C]. Computer Graphics(SIGGRAPH′93 Conference Proceedings), July 1993,pp,279-288.
[5]A.Lippman. Movie-Maps:an Application of the Optical Videodisc to Computer Graphics[C].Computer Graphics, ACM SIGGRAPH, 1980,(8):32-42.
[6]E.H.Adelson, J.Bergeb. The plenoptic function and the elements of early vision[J]. Computational Models of Visual Processing. 1991:3-20.
[7]L.McMillan, G.Bishop. Plenoptic modeling:An Image-based Rendering System[C]. Computer Graphics Proceedings, Annual Conference Series, ACM SIGGRAPH, 1995:39-46.
[8]M.Levoy and P.Hanrahan. Light Field Rendering[C]. Computer Graphics, ACM SIGGRAPH′96, 1996(8):31-42.
[9]S.J.Gortler, R.Gzeszczuk, R.Szeliski, M.F.Coen. The Lumigraph[C]. Computer Graphics, ACM SIGGRAPH′96,1996,(8):43-54.
[10]H.Y.Shum, L.W.He. Rendering with concentric mosaics[C]. In Proc.SIGGRAPH′99, 1999:299-306.
[11]H-Y Shum, R.Szeliski. Construction and Refinement of Panoamic Mosaics with Global and Local Alignment[C]. Sixth International Conference on Computer Vision(ICCV′98), Bombay, 1998(1):953-958.
[12]S.E.Chen, L.Williams. View Interpolation for Image Synthesis[C]. Computer Graphics, ACM SIGGRAPH′93,1993,(8):279-288.
[13]Wolberg G. Digital Image Warping[J]. IEEE Computer Society Press,1990.
[14]Hartley R.I. In Defense of the 8-Point algorithm[C]. In Fifth International Conf. On Computer Vision(ICCV′95),1064-1070, Cambridge, Massachusetts, June 1995.
[15]Hartley R.I. In Defense of the Eight-Point Algorithm[M]. IEEE Trans.On Pattern Analysis and Machine Intelligence Vol.19,NO.6,581-593,1997.
[16]Laveau S. and Faugeas O. 3-D scene representation as a collection of images and fundamental matries[J]. Technical Report 2205, INRIA-Sohia, Antipolis, Feb.1994.
[17]Faugeas O. and Laveau S. et al. 3-D Reconstruction of Urban Scenes from Sequences of Images[J]. Technical Report 2572, INRIA-Sophia, Antipolis, June.1995.
[18]Ying X, Hu I, Zha H. Fisheye lenses calibration using straight-line spherical perspective constraint[C]// ACCV,2006,Proc. Of 7th Asian conf. on computer Vision. India:Hyderabad,ACCV,2006:591-600.
[19]英向华,胡占义.一种基于球面投影约束的鱼眼镜头校正方法[J].计算机学报,2003,26(12):1702-1708.
[20]邱志强,陆宏伟,于起峰.用射影不变性纠正鱼眼镜头畸变[J].应用光学, 2003:24(5):36-38.
[21]朱正明.基于特征点匹配的图像匹配系统研究[D].学位论文.北京:北京大学,2002.
[22]John A R, Xiu ping J. Remote Sensing Digital Image Ana1ysis[J]. Springer-Verlag BerlinHeidelberg,1998.
[23]Bhosle Udhav, Chaudhuri Subhasis, Roy Sumantra Dutta. A fast method for image mosaicing using Geometric Hashing[J]. IEEE Journal of Research,2002,48(3-4):317-324.
[24]Satyan C, Seth T. Spherical Mosaics with Quaternions and Dense Correlation[J]. International Journal of Computer Vision, 2000,37(3):259-273.
[25]Kuglin C D, Hines D C. The phase correlation image alignment method[C]. IEEE Proc of IEEE International Conference on Cybernetics and Society. New York:IEEE,1975.163-165.
[26]Reddy B S, Chatterji B N. An FFT-based Technique for Translation, Rotation and Scale-invariant Image Registration[J]。IEEE Trans. Image Processing,1996,5(8):1266-1271.
[27]Xie Hongjie, et al. Automatic image registration based on FFT algorithm and IDL/ENVI[C]. In:Intl. Conf. on Remote Sensing and GIS/GPS-ICORG,2000.
[28]Badra F, Qumsieh A, Dudek G. Rotation and Zooming in Image Mosaicing[J].In IEEE WAVCV,1998:50-55.
[29]Satyan C, Seth T. Spherical Mosaics with quaternions and Dense Correlation[J]. International Journal of Computer Vision,2000,37(3):259-273.
[30]Yi hong G, Guido P, David L. A Robust Image Mosaicing Technique Capable of Creating Integrated Panoramas[C]. Proceedings of IEEE International Conference on Information Visualization,Los Alamitos, CA,USA,1999,24-29.
[31]Szeliski R. Video Mosaics for Virtual Enviroments[J]. IEEE Computer Graphics and Applications, 1996, 16(2):22-30.
[32]P.J.Burt, E.H.Adelson. A Multiresolution Spline With Application to Image Mosaics[J]. ACM Transactions on Graphics. 1983,2(4):217-236.
[33]K .Miyamoto. Fish eye lens[J]. Jour.Opt.Soc.Amer.,54(8)1974,1069.
[34]戴建宁.鱼眼镜头的成像理论与优化设计[D].博士学位论文.长沙:国防科学技术大学,1999.
[35]Maybank S. Theory of Reconstruction from Image Motion[C]. In: Springer Series in Information Sciences, New York:Springer-Verlag,1993.
[36]英向华.全向摄像机标定技术研究[D].博士学位论文.沈阳:中国科学院自动化研究所,2004.
[37]Swaminathan R,Nayar S K. Non-metric calibration of wid-angle lenses and polycameras[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence,2000,22(10):1172-1178.
[38]Kang S B. Radial distortion snakes[J]. In:Proceedings of IAPR Workshop on Machine Vision Applications,Tokyo,2000.603-606.
[39]Devernay F, Faugeras O. Straight lines have to be straight:Automatic calibration and removal of distortion from scenes of structured environments[J]. Machine Vision and Applications,2001,13(1):14-24.
[40]Swaminathan R, Nayar S K. Non-metric calibration of wid-angle lenses and polycameras[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(10):1172-1178.
[41]Xiong Y, Turkowski K. Creating image-based VR using a self-calibrating fisheye lens[C]. In:Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Puerto Rico,1997,237-243.
[42]赵弘,等.基于Levenberg-Marquardt算法的神经网络监督控制[J].西安交通大学学报,2002(5).
[43]李弼程,等.智能图像处理技术[M].北京:电子工业出版社,2004,7.
[44]章毓晋.图像理解与计算机视觉[M].北京:清华大学出版社,2000.
[45]Gonzalez R C,Wintz P. 1987. Digital Image Processing[M].2nd. ed. Addison-Wesley.
[46]Zhang Y J. 1992 a. Improving the accuracym of direct histogram specification[J]. IEE Electronics Letters,28:213-214.
[47]Kenneth.R.Castleman著,朱志刚等译.数字图像处理[M].北京:电子工业出版社,1998,9.
[48]张志佳.图像特征提取及目标匹配的理论和方法研究[D].博士学位论文.沈阳:中国科学院自动化研究所,2005.
[49]卜彦龙.单通道红外寻的防空导弹电子稳像技术研究[D].硕士学位论文.长沙:国防科学技术大学,2004.
[50]叶耘恺.基于边缘特征的图像配准方法研究.硕士学位论文[D].重庆:重庆大学,2004.
[51]贾天旭等.基于Bubble小波的多尺度边缘提取[J].电子学报,1996.Vol.24.No.4:117-121.
[52]魏海等.反对称双正交小波应用于多尺度边缘提取的研究[J].电子学报,2002.Vo1.30.No.3:313-320.
[53]贾天旭等.基于一类中心B样条二进小波的多尺度边缘提取[J].西安交通大学学报,1996.Vol.30.No.l:91-113.
[54]徐琼.边缘检测梯度法的小波实现[J].四川轻化工学院学报, 1998,Vo1.24.No.2:11-14.
[55]解梅,顾德仁.使用小波变换的图像边缘检测算法[J].电子科技大学学报,1996,Vo1.25.No.4:353-356.
[56]ROSENFELD A. WESZKA J S. An improved method of angel detection on digital curves[J]. IEEE Trans. Comput,1975,C-24(9),940-941.
[57]FREEMAN H. DAVIS L S. A corner finding algorithm for chain-coded curves[J]. IEEE Trans Computers,1977,26(3):297-303.
[58]乔宇,黄席樾,柴毅,周欣.基于自适应直线拟合的角点检测[J].重庆大学学报.Vol.26,No.2,2003.
[59]Hartley R, Zisserman A.Multiple view geometry in computer vision[M].London:2nd Cambridge University Press,2004:8-11.
[60]智平华,孙季丰.真实场景的图像拼接方法[J].中国图像图形学报, 2003,(8):377-380.
[61]李寒,牛纪桢,郭禾.基于特征点的全自动无缝图像拼接方法[J].计算机工程与设计,2007,Vol.28,No.9.
[2]张茂军.虚拟现实系统[M].北京:科学出版社,2001.
[3]杨东录.犯罪现场勘查实用技术[M].西安:陕西人民出版社,2003.
[4]S.E.Chen, L.Williams. View Interpolation for Image Synthesis[C]. Computer Graphics(SIGGRAPH′93 Conference Proceedings), July 1993,pp,279-288.
[5]A.Lippman. Movie-Maps:an Application of the Optical Videodisc to Computer Graphics[C].Computer Graphics, ACM SIGGRAPH, 1980,(8):32-42.
[6]E.H.Adelson, J.Bergeb. The plenoptic function and the elements of early vision[J]. Computational Models of Visual Processing. 1991:3-20.
[7]L.McMillan, G.Bishop. Plenoptic modeling:An Image-based Rendering System[C]. Computer Graphics Proceedings, Annual Conference Series, ACM SIGGRAPH, 1995:39-46.
[8]M.Levoy and P.Hanrahan. Light Field Rendering[C]. Computer Graphics, ACM SIGGRAPH′96, 1996(8):31-42.
[9]S.J.Gortler, R.Gzeszczuk, R.Szeliski, M.F.Coen. The Lumigraph[C]. Computer Graphics, ACM SIGGRAPH′96,1996,(8):43-54.
[10]H.Y.Shum, L.W.He. Rendering with concentric mosaics[C]. In Proc.SIGGRAPH′99, 1999:299-306.
[11]H-Y Shum, R.Szeliski. Construction and Refinement of Panoamic Mosaics with Global and Local Alignment[C]. Sixth International Conference on Computer Vision(ICCV′98), Bombay, 1998(1):953-958.
[12]S.E.Chen, L.Williams. View Interpolation for Image Synthesis[C]. Computer Graphics, ACM SIGGRAPH′93,1993,(8):279-288.
[13]Wolberg G. Digital Image Warping[J]. IEEE Computer Society Press,1990.
[14]Hartley R.I. In Defense of the 8-Point algorithm[C]. In Fifth International Conf. On Computer Vision(ICCV′95),1064-1070, Cambridge, Massachusetts, June 1995.
[15]Hartley R.I. In Defense of the Eight-Point Algorithm[M]. IEEE Trans.On Pattern Analysis and Machine Intelligence Vol.19,NO.6,581-593,1997.
[16]Laveau S. and Faugeas O. 3-D scene representation as a collection of images and fundamental matries[J]. Technical Report 2205, INRIA-Sohia, Antipolis, Feb.1994.
[17]Faugeas O. and Laveau S. et al. 3-D Reconstruction of Urban Scenes from Sequences of Images[J]. Technical Report 2572, INRIA-Sophia, Antipolis, June.1995.
[18]Ying X, Hu I, Zha H. Fisheye lenses calibration using straight-line spherical perspective constraint[C]// ACCV,2006,Proc. Of 7th Asian conf. on computer Vision. India:Hyderabad,ACCV,2006:591-600.
[19]英向华,胡占义.一种基于球面投影约束的鱼眼镜头校正方法[J].计算机学报,2003,26(12):1702-1708.
[20]邱志强,陆宏伟,于起峰.用射影不变性纠正鱼眼镜头畸变[J].应用光学, 2003:24(5):36-38.
[21]朱正明.基于特征点匹配的图像匹配系统研究[D].学位论文.北京:北京大学,2002.
[22]John A R, Xiu ping J. Remote Sensing Digital Image Ana1ysis[J]. Springer-Verlag BerlinHeidelberg,1998.
[23]Bhosle Udhav, Chaudhuri Subhasis, Roy Sumantra Dutta. A fast method for image mosaicing using Geometric Hashing[J]. IEEE Journal of Research,2002,48(3-4):317-324.
[24]Satyan C, Seth T. Spherical Mosaics with Quaternions and Dense Correlation[J]. International Journal of Computer Vision, 2000,37(3):259-273.
[25]Kuglin C D, Hines D C. The phase correlation image alignment method[C]. IEEE Proc of IEEE International Conference on Cybernetics and Society. New York:IEEE,1975.163-165.
[26]Reddy B S, Chatterji B N. An FFT-based Technique for Translation, Rotation and Scale-invariant Image Registration[J]。IEEE Trans. Image Processing,1996,5(8):1266-1271.
[27]Xie Hongjie, et al. Automatic image registration based on FFT algorithm and IDL/ENVI[C]. In:Intl. Conf. on Remote Sensing and GIS/GPS-ICORG,2000.
[28]Badra F, Qumsieh A, Dudek G. Rotation and Zooming in Image Mosaicing[J].In IEEE WAVCV,1998:50-55.
[29]Satyan C, Seth T. Spherical Mosaics with quaternions and Dense Correlation[J]. International Journal of Computer Vision,2000,37(3):259-273.
[30]Yi hong G, Guido P, David L. A Robust Image Mosaicing Technique Capable of Creating Integrated Panoramas[C]. Proceedings of IEEE International Conference on Information Visualization,Los Alamitos, CA,USA,1999,24-29.
[31]Szeliski R. Video Mosaics for Virtual Enviroments[J]. IEEE Computer Graphics and Applications, 1996, 16(2):22-30.
[32]P.J.Burt, E.H.Adelson. A Multiresolution Spline With Application to Image Mosaics[J]. ACM Transactions on Graphics. 1983,2(4):217-236.
[33]K .Miyamoto. Fish eye lens[J]. Jour.Opt.Soc.Amer.,54(8)1974,1069.
[34]戴建宁.鱼眼镜头的成像理论与优化设计[D].博士学位论文.长沙:国防科学技术大学,1999.
[35]Maybank S. Theory of Reconstruction from Image Motion[C]. In: Springer Series in Information Sciences, New York:Springer-Verlag,1993.
[36]英向华.全向摄像机标定技术研究[D].博士学位论文.沈阳:中国科学院自动化研究所,2004.
[37]Swaminathan R,Nayar S K. Non-metric calibration of wid-angle lenses and polycameras[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence,2000,22(10):1172-1178.
[38]Kang S B. Radial distortion snakes[J]. In:Proceedings of IAPR Workshop on Machine Vision Applications,Tokyo,2000.603-606.
[39]Devernay F, Faugeras O. Straight lines have to be straight:Automatic calibration and removal of distortion from scenes of structured environments[J]. Machine Vision and Applications,2001,13(1):14-24.
[40]Swaminathan R, Nayar S K. Non-metric calibration of wid-angle lenses and polycameras[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(10):1172-1178.
[41]Xiong Y, Turkowski K. Creating image-based VR using a self-calibrating fisheye lens[C]. In:Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Puerto Rico,1997,237-243.
[42]赵弘,等.基于Levenberg-Marquardt算法的神经网络监督控制[J].西安交通大学学报,2002(5).
[43]李弼程,等.智能图像处理技术[M].北京:电子工业出版社,2004,7.
[44]章毓晋.图像理解与计算机视觉[M].北京:清华大学出版社,2000.
[45]Gonzalez R C,Wintz P. 1987. Digital Image Processing[M].2nd. ed. Addison-Wesley.
[46]Zhang Y J. 1992 a. Improving the accuracym of direct histogram specification[J]. IEE Electronics Letters,28:213-214.
[47]Kenneth.R.Castleman著,朱志刚等译.数字图像处理[M].北京:电子工业出版社,1998,9.
[48]张志佳.图像特征提取及目标匹配的理论和方法研究[D].博士学位论文.沈阳:中国科学院自动化研究所,2005.
[49]卜彦龙.单通道红外寻的防空导弹电子稳像技术研究[D].硕士学位论文.长沙:国防科学技术大学,2004.
[50]叶耘恺.基于边缘特征的图像配准方法研究.硕士学位论文[D].重庆:重庆大学,2004.
[51]贾天旭等.基于Bubble小波的多尺度边缘提取[J].电子学报,1996.Vol.24.No.4:117-121.
[52]魏海等.反对称双正交小波应用于多尺度边缘提取的研究[J].电子学报,2002.Vo1.30.No.3:313-320.
[53]贾天旭等.基于一类中心B样条二进小波的多尺度边缘提取[J].西安交通大学学报,1996.Vol.30.No.l:91-113.
[54]徐琼.边缘检测梯度法的小波实现[J].四川轻化工学院学报, 1998,Vo1.24.No.2:11-14.
[55]解梅,顾德仁.使用小波变换的图像边缘检测算法[J].电子科技大学学报,1996,Vo1.25.No.4:353-356.
[56]ROSENFELD A. WESZKA J S. An improved method of angel detection on digital curves[J]. IEEE Trans. Comput,1975,C-24(9),940-941.
[57]FREEMAN H. DAVIS L S. A corner finding algorithm for chain-coded curves[J]. IEEE Trans Computers,1977,26(3):297-303.
[58]乔宇,黄席樾,柴毅,周欣.基于自适应直线拟合的角点检测[J].重庆大学学报.Vol.26,No.2,2003.
[59]Hartley R, Zisserman A.Multiple view geometry in computer vision[M].London:2nd Cambridge University Press,2004:8-11.
[60]智平华,孙季丰.真实场景的图像拼接方法[J].中国图像图形学报, 2003,(8):377-380.
[61]李寒,牛纪桢,郭禾.基于特征点的全自动无缝图像拼接方法[J].计算机工程与设计,2007,Vol.28,No.9.