三维鞋底花纹的重构和匹配
|
摘要
随着犯罪形式的变化,在案件侦查过程中,足迹检验技术得到了越来越广泛的应用,也越来越得到司法部门和公安机关的重视。鞋印图像作为一个包含着丰富信息的高冗余度载体,能较好地反映出案发现场罪犯所留下的犯罪痕迹。然而,现场拍照所取得的图像有很多信息是残缺的,并且很难以让肉眼识别,因此,如何使图像信息得到最大限度的使用是一个迫在眉睫的问题。另外,随着罪犯的反侦察能力的不断提升,已经成熟的2D技术逐渐有了其局限性,3D图像处理得到了更为广泛的使用。
本文提出构造一个对犯罪现场提取的鞋印进行精确检测的系统。利用激光扫描仪对现场采集的鞋印进行数据提取。所得到的数据为三维点云数据,在MFC平台下,利用OpenGL软件进行鞋印图像的点云显示。标记包围盒算法被用于对数据进行约简和重构,同时利用ICP算法对现场采集数据和数据库中的鞋印数据进行比对和匹配,得出现场采集鞋印的型号及相关信息,从而为公安人员破案提供有效的参考和依据。
With the change of means of crime, footprint detecting technology has been found more and more applications in investigating and solve the criminal cases, and it has been more concerned by justice and public security departments. As an information carrier, of high redundancy, footprint images can contain plenty of information. Therefore, shoeprint images can indicate trace of crime quite well in crime scene. However, the pictures taken in crime scene are often incomplete and, as a result, are quite difficult to recognize. Therefore, how to take full advantage of shoeprint images is an urgent issue nowadays. Moreover, with enhance of criminals' anti- detection capability, the state-of-the-art 2D graphic processing technology has shown its increasing limitations. Consequently,3D graphic processing attracts more intention of researchers.
This dissertation presents a precise shoeprint detection system for footprints taken from the crime scene. The principle of the system is that, firstly, the footprint data is extracted the footprint pictures taken from in crime scene by laser scanner, the collected data is stored the form of 3D point cloud. Then the OpenGL software is used to visualize the point cloud of footprint in the platform of MFC. Next, in order to get the model and some relevant information of the footprint, it is necessary to do the deta reduction and reconstruction by using the marked bounding box method. Afterwards, the ICP algorithm is used to compare and match the footprint data collected from crime scene and the footprint data in database, in order to get the information such as the size and the manufacturer of the shoes. In conclusion, this system is much helpful for police.
本文提出构造一个对犯罪现场提取的鞋印进行精确检测的系统。利用激光扫描仪对现场采集的鞋印进行数据提取。所得到的数据为三维点云数据,在MFC平台下,利用OpenGL软件进行鞋印图像的点云显示。标记包围盒算法被用于对数据进行约简和重构,同时利用ICP算法对现场采集数据和数据库中的鞋印数据进行比对和匹配,得出现场采集鞋印的型号及相关信息,从而为公安人员破案提供有效的参考和依据。
With the change of means of crime, footprint detecting technology has been found more and more applications in investigating and solve the criminal cases, and it has been more concerned by justice and public security departments. As an information carrier, of high redundancy, footprint images can contain plenty of information. Therefore, shoeprint images can indicate trace of crime quite well in crime scene. However, the pictures taken in crime scene are often incomplete and, as a result, are quite difficult to recognize. Therefore, how to take full advantage of shoeprint images is an urgent issue nowadays. Moreover, with enhance of criminals' anti- detection capability, the state-of-the-art 2D graphic processing technology has shown its increasing limitations. Consequently,3D graphic processing attracts more intention of researchers.
This dissertation presents a precise shoeprint detection system for footprints taken from the crime scene. The principle of the system is that, firstly, the footprint data is extracted the footprint pictures taken from in crime scene by laser scanner, the collected data is stored the form of 3D point cloud. Then the OpenGL software is used to visualize the point cloud of footprint in the platform of MFC. Next, in order to get the model and some relevant information of the footprint, it is necessary to do the deta reduction and reconstruction by using the marked bounding box method. Afterwards, the ICP algorithm is used to compare and match the footprint data collected from crime scene and the footprint data in database, in order to get the information such as the size and the manufacturer of the shoes. In conclusion, this system is much helpful for police.
引文
[1]王磊.浅谈足迹在案件侦查中的作用[J].法制与社会,2008.
[2]史力民,毕胜,姜吉.足迹信息的计算机管理与查询[J].刑事技术,2004.
[3]姜维珉,孔正方,吴志军.足迹分析与比对系统研发思路[J].中国刑事警察,2003.
[4]Patil P M, Kulkarni J V. Rotation and intensity invariant shoeprint matching using Gabor transform with application to forensic science [J]. Pattern Recognition,2004,42(7): 1308-1317.
[5]Geradts Z, Bjihold J. New developments in forensic image processing and pattern recognition [J]. Journal of the Forensic Science Society,2001,41(3):159-166.
[6]白雪.足迹比对算法的研究与实现[D].中国农业大学,2007.
[7]R.Buss著.3D计算机图形学(OpenGL版)[M].唐龙等译.北京：清华大学出版社,2006.
[8]Shreiner D, Woo M, Neider J等著.OpenGL编程指南[M].徐波译.北京：机械工业出版社,2008.
[9]陈太喜,方亮,刘艳伟等.基于OpenGL的STL数据模型建模方法[J].机床与液压,2009,37(3)：155-158.
[10]周莉,苏鸿根.通用3D模型文件格式和算法的研究及其OpenGL实现[J].计算机工程与设计,2009,30(2)：433-439.
[11]陈青峰,谌永祥,李永桥.基于Visual C++的点云数据预处理的设计研究[J].机械工程与自动化,2009,1(2)：26-30.
[12]许智钦,闰明,张宝峰.逆向工程技术三维激光扫描测量[J].天津大学学报,2001,1(3)：404-408.
[13]孙鑫,余安萍.VC++深入详解[M].北京：电子工业出版社,2006.
[14]吴爱兰,楼建列.基于OpenGL的三维图形的绘制[J].信息科技,2007：122-124.
[15]刘敏莺,黄文骞.基于OpenGL的地形三维可视化[J].海洋测绘,2002：26-29.
[16]马婷.点云数据的显示与配准[D].吉林大学,2007.
[17]曾敬文,朱照荣,丁锐.基于立方体网格的数据点云约简和体积计算方法[J].测绘科学,2008,33(6)：81-82.
[18]阎树田,惠相君,程萍.逆向工程中密集数据采样以及网格化技术[J].中国工程机械报,2006,32(6)：44-46.
[19]洪军,丁玉成,曹亮.逆向工程中的测量数据精简技术研究[J].西安交通大学学报,2004,38(7)：661-664.
[20]刘涛,徐铮,沙成梅等.基于包围盒法的散乱点云数据的曲率精简[J].科学技术与工程,2009,9(12)：3333-3336.
[21]孙肖霞,孙殿柱.反向工程中测量数据的精简算法[J].机械设计与制造,2006,12(2)：37-38.
[22]程晨.浅析散乱点云数据的三角网格面重构[J].科技信息,2009：794-800.
[23]Dey T K, Giesen J and Hudson J. Decimating samples for mesh simplification [C]. Proceedings of 13th Canadian Conference on Computational Geometry, Waterloo, Canada, 2001:85-88.
[24]Pauly M, Gross M and Kobbelt L P. Efficient simplification of point-sampled surfaces [C]. Proceedings of 13th IEEE Visualization Conf., Boston, USA,2002.
[25]Linsen L. Point cloud representation [C]. Proceedings of CS Technical Report 2001-3, Universitat Karlsruhe, Germany,2001.
[26]Lorensen W E, Cline H E. Marching cubes:A high resolution 3D surface construction algorithm [J]. Computer Graphics,1987,21(4):163-169.
[27]Wang X Y, Wu J H, Zhao Q M. A Maeked Bouding Box Method for Image Data Reduction and Reconstruction of Sole Patterns [C]. Proceedings of International Conference on Software and Computing Technology, Kunming, China.2010:11-14.
[28]何炳蔚.逆向工程中密集数据采样以及网格化技术[J].中国工程机械学报,2006,4(4)：474-482.
[29]周会成,陈吉红,周云飞,等.三维散乱点云的分割与几何形体重构[J].机械与电子,2003,1(6)：3-6.
[30]Wang Y B, Sheng Y H, Lu G N. Feature-constrained surface reconstruction method for point cloud data acquired by 3d laser scanner [C]. Proceedings of SPIE, Photonics'2008: Optical and Digital Image Processing. Strasbourg, France,2008:1-12.
[31]Sankaranarayanan J, Samet H, Varshney A, A fast all nearest neighbor algorithm for applications involving large point-clouds [J]. Computers & Graphics.2007:157-174.
[32]Woo H, Kang E, Wang S Y. A new segmentation method for point cloud data [J]. International Journal of Machine Tools & manufacture,2002:167-178.
[33]Li C M, Li Z X, Wang Y H. Some improvements of the marching cubes algorithm for the rendering of an ore body [J]. J. China Univ. Mining & Technol.,2008:194-198.
[34]Zhang L Y, Zhou R R, Zhou L S. Model reconstruction from cloud data [J]. Journal of Materials Processing Technology.2003:494-498.
[35]Mahmoudi M, Sapiro G. Three-dimensional point cloud recognition via distributions of geometric distances [J]. Graphical Models.2009:22-31.
[36]Lin H W, Tai C L, Wang G J. A mesh reconstruction algorithm driven by an intrinsic property of a point cloud [J]. Computer-Aided Design,2004:1-9.
[37]Liu B, Shangguan N, Jiang K Y. Triangular Mesh Model Reconstruction from Scan Point Clouds Based on Template [J]. Tsinghua Science and Technology,2009,14(2):56-61.
[38]Ouyang D S, Feng H Y. On the normal vector estimation for point cloud data from smooth surfaces [J]. Computer-Aided Design.2005:1071-1079.
[39]鄢腊梅,袁友伟,周锋.基于点云数据重构三维网格模型的神经网络法[J].机械设计,2006,23(4)：14-15.
[40]韦争亮,钟约先,袁朝龙.散乱三维测量数据构建三角形网格模型的研究[J].机械设计与研究,2008,24(1)：61-63.
[41]黄芳,史振华,张峰.面向心脏表面断层数据的三角网格重构方法[J].计算机系统应用,2010,19(9)：92-96.
[42]林志浩,姜晓峰,王鹏远.截面数据重构三角网格算法的研究[J].微电子学与计算机,2007,24(3)：130-132.
[43]张政.点云数据配准算法研究[D].山东大学,2008.
[44]吴禄慎,孔维敬.基于特征点的改进ICP三维点云配准技术[J].南昌大学学报工科版,2008,30(3)：294-297.
[45]张旭,王青,柯映林.基于统计特征的点云模型匹配技术[J].计算机集成制造系统,2008,14(8)：1539-1544.
[46]张凯.三维激光扫描数据的空间配准研究[D].南京师范大学,2008.
[47]戴静兰.海量点云预处理算法研究[D].浙江大学,2006.
[48]刘宇,熊有伦.基于有界k-d树的最近点搜索算法[J].华中科技大学学报(自然科学版).2008,36(7)：73-76.
[49]高珊珊.基于三维激光扫描仪的点云配准[D].南京理工大学,2008.
[50]张曼,沈旭昆.一种基于尺度空间的三维点云数据配准算法[J].系统仿真学报.2009,21(1)：131-135.
[51]Matusik W, Pfister H, Zieglar R. Acquisition and rendering of transparent and refractive objects [C]. Proceedings of Thirteenth Eurographics Workshop on Rendering, Pisa, Italy. 2002:267-278.
[52]陈春萍.基于OpenGL的隧道掌字面三维重建系统的设计与实现[D].西南交通大学,2005.
[2]史力民,毕胜,姜吉.足迹信息的计算机管理与查询[J].刑事技术,2004.
[3]姜维珉,孔正方,吴志军.足迹分析与比对系统研发思路[J].中国刑事警察,2003.
[4]Patil P M, Kulkarni J V. Rotation and intensity invariant shoeprint matching using Gabor transform with application to forensic science [J]. Pattern Recognition,2004,42(7): 1308-1317.
[5]Geradts Z, Bjihold J. New developments in forensic image processing and pattern recognition [J]. Journal of the Forensic Science Society,2001,41(3):159-166.
[6]白雪.足迹比对算法的研究与实现[D].中国农业大学,2007.
[7]R.Buss著.3D计算机图形学(OpenGL版)[M].唐龙等译.北京：清华大学出版社,2006.
[8]Shreiner D, Woo M, Neider J等著.OpenGL编程指南[M].徐波译.北京：机械工业出版社,2008.
[9]陈太喜,方亮,刘艳伟等.基于OpenGL的STL数据模型建模方法[J].机床与液压,2009,37(3)：155-158.
[10]周莉,苏鸿根.通用3D模型文件格式和算法的研究及其OpenGL实现[J].计算机工程与设计,2009,30(2)：433-439.
[11]陈青峰,谌永祥,李永桥.基于Visual C++的点云数据预处理的设计研究[J].机械工程与自动化,2009,1(2)：26-30.
[12]许智钦,闰明,张宝峰.逆向工程技术三维激光扫描测量[J].天津大学学报,2001,1(3)：404-408.
[13]孙鑫,余安萍.VC++深入详解[M].北京：电子工业出版社,2006.
[14]吴爱兰,楼建列.基于OpenGL的三维图形的绘制[J].信息科技,2007：122-124.
[15]刘敏莺,黄文骞.基于OpenGL的地形三维可视化[J].海洋测绘,2002：26-29.
[16]马婷.点云数据的显示与配准[D].吉林大学,2007.
[17]曾敬文,朱照荣,丁锐.基于立方体网格的数据点云约简和体积计算方法[J].测绘科学,2008,33(6)：81-82.
[18]阎树田,惠相君,程萍.逆向工程中密集数据采样以及网格化技术[J].中国工程机械报,2006,32(6)：44-46.
[19]洪军,丁玉成,曹亮.逆向工程中的测量数据精简技术研究[J].西安交通大学学报,2004,38(7)：661-664.
[20]刘涛,徐铮,沙成梅等.基于包围盒法的散乱点云数据的曲率精简[J].科学技术与工程,2009,9(12)：3333-3336.
[21]孙肖霞,孙殿柱.反向工程中测量数据的精简算法[J].机械设计与制造,2006,12(2)：37-38.
[22]程晨.浅析散乱点云数据的三角网格面重构[J].科技信息,2009：794-800.
[23]Dey T K, Giesen J and Hudson J. Decimating samples for mesh simplification [C]. Proceedings of 13th Canadian Conference on Computational Geometry, Waterloo, Canada, 2001:85-88.
[24]Pauly M, Gross M and Kobbelt L P. Efficient simplification of point-sampled surfaces [C]. Proceedings of 13th IEEE Visualization Conf., Boston, USA,2002.
[25]Linsen L. Point cloud representation [C]. Proceedings of CS Technical Report 2001-3, Universitat Karlsruhe, Germany,2001.
[26]Lorensen W E, Cline H E. Marching cubes:A high resolution 3D surface construction algorithm [J]. Computer Graphics,1987,21(4):163-169.
[27]Wang X Y, Wu J H, Zhao Q M. A Maeked Bouding Box Method for Image Data Reduction and Reconstruction of Sole Patterns [C]. Proceedings of International Conference on Software and Computing Technology, Kunming, China.2010:11-14.
[28]何炳蔚.逆向工程中密集数据采样以及网格化技术[J].中国工程机械学报,2006,4(4)：474-482.
[29]周会成,陈吉红,周云飞,等.三维散乱点云的分割与几何形体重构[J].机械与电子,2003,1(6)：3-6.
[30]Wang Y B, Sheng Y H, Lu G N. Feature-constrained surface reconstruction method for point cloud data acquired by 3d laser scanner [C]. Proceedings of SPIE, Photonics'2008: Optical and Digital Image Processing. Strasbourg, France,2008:1-12.
[31]Sankaranarayanan J, Samet H, Varshney A, A fast all nearest neighbor algorithm for applications involving large point-clouds [J]. Computers & Graphics.2007:157-174.
[32]Woo H, Kang E, Wang S Y. A new segmentation method for point cloud data [J]. International Journal of Machine Tools & manufacture,2002:167-178.
[33]Li C M, Li Z X, Wang Y H. Some improvements of the marching cubes algorithm for the rendering of an ore body [J]. J. China Univ. Mining & Technol.,2008:194-198.
[34]Zhang L Y, Zhou R R, Zhou L S. Model reconstruction from cloud data [J]. Journal of Materials Processing Technology.2003:494-498.
[35]Mahmoudi M, Sapiro G. Three-dimensional point cloud recognition via distributions of geometric distances [J]. Graphical Models.2009:22-31.
[36]Lin H W, Tai C L, Wang G J. A mesh reconstruction algorithm driven by an intrinsic property of a point cloud [J]. Computer-Aided Design,2004:1-9.
[37]Liu B, Shangguan N, Jiang K Y. Triangular Mesh Model Reconstruction from Scan Point Clouds Based on Template [J]. Tsinghua Science and Technology,2009,14(2):56-61.
[38]Ouyang D S, Feng H Y. On the normal vector estimation for point cloud data from smooth surfaces [J]. Computer-Aided Design.2005:1071-1079.
[39]鄢腊梅,袁友伟,周锋.基于点云数据重构三维网格模型的神经网络法[J].机械设计,2006,23(4)：14-15.
[40]韦争亮,钟约先,袁朝龙.散乱三维测量数据构建三角形网格模型的研究[J].机械设计与研究,2008,24(1)：61-63.
[41]黄芳,史振华,张峰.面向心脏表面断层数据的三角网格重构方法[J].计算机系统应用,2010,19(9)：92-96.
[42]林志浩,姜晓峰,王鹏远.截面数据重构三角网格算法的研究[J].微电子学与计算机,2007,24(3)：130-132.
[43]张政.点云数据配准算法研究[D].山东大学,2008.
[44]吴禄慎,孔维敬.基于特征点的改进ICP三维点云配准技术[J].南昌大学学报工科版,2008,30(3)：294-297.
[45]张旭,王青,柯映林.基于统计特征的点云模型匹配技术[J].计算机集成制造系统,2008,14(8)：1539-1544.
[46]张凯.三维激光扫描数据的空间配准研究[D].南京师范大学,2008.
[47]戴静兰.海量点云预处理算法研究[D].浙江大学,2006.
[48]刘宇,熊有伦.基于有界k-d树的最近点搜索算法[J].华中科技大学学报(自然科学版).2008,36(7)：73-76.
[49]高珊珊.基于三维激光扫描仪的点云配准[D].南京理工大学,2008.
[50]张曼,沈旭昆.一种基于尺度空间的三维点云数据配准算法[J].系统仿真学报.2009,21(1)：131-135.
[51]Matusik W, Pfister H, Zieglar R. Acquisition and rendering of transparent and refractive objects [C]. Proceedings of Thirteenth Eurographics Workshop on Rendering, Pisa, Italy. 2002:267-278.
[52]陈春萍.基于OpenGL的隧道掌字面三维重建系统的设计与实现[D].西南交通大学,2005.