通航船舶三维吃水动态检测技术的研究
|
摘要
我国内河水运资源丰富,内河航运飞速发展,有力地推动了我国国民经济的发展。随着内河航运量的不断增大,内河船舶超吃水违规通航的现象也日益增多,严重影响了内河航道的通航安全和通航效率,阻碍了航运经济的健康发展。然而,相关执法部门仍只能采取登船检测船舶吃水线的方式查处船舶超吃水的违规现象,既耗费时间又影响了船舶的正常通航。目前尚无有效的技术手段进行快速的离船式通航船舶吃水测量。
检测船舶的三维吃水不仅能够获得船舶的最大吃水值,也可以获取船舶三维吃水数据模型等大量信息,是检测船舶超载以及检测船底形变的数据基础。根据通航船舶三维吃水的检测要求,研制了基于单波束超声波传感器阵列的通航船舶三维吃水动态检测系统,在保证通航效率的前提下,实现了对并行通航船舶三维吃水的快速离船式动态检测,并对超吃水船舶进行预警。目前系统已在长江三峡船闸投入使用,运行状态良好。
检测系统的超声波传感器阵列安装于检测支架上,采用检测门同步自动升降子系统将检测支架同步、平稳地放置于水下合适的测量位置,避免影响船舶正常通航。船舶三维吃水数据采集和处理子系统通过传感器阵列获得船舶三维吃水数据,进行异常数据剔除、多船吃水数据分割以及三维吃水数据模型修正处理,将数据处理后的吃水数据以三维形式显示于计算机上,并计算出每艘被测船舶对应的最大吃水,用无线网络发送至信息综合管理系统进行超吃水预警。
利用通航船舶三维吃水动态检测系统,获取船舶原始的三维吃水数据模型,其中存在较多的异常数据并且混合了多艘被测船舶的吃水数据。对于原始的三维吃水数据模型,采用点云数据中值滤波算法剔除“类随机噪声”,采用最小二乘法直线回归预测法剔除“船舶尾流噪声”。对于预处理后的船舶三维吃水数据,本文分析了船体不同部位的吃水数据特征,并结合系统的动态检测要求,提出了基于数据特提取的点云数据动态分割算法,对三维吃水数据模型进行动态分割,分离出并行通航船舶各自的三维吃水数据模型,并采用系统实际测量数据,对整体算法进行了验证。同时,本文介绍了基于OpenGL的船舶吃水数据动态三维显示技术,并采用系统实测数据进行了功能验证。
为了进一步提高系统的检测精度,本文提出采用船舶三维吃水模型匹配技术实现船舶底部形变检测以及船舶三维吃水模型的修正。本文详细介绍了船舶三维吃水模型匹配技术的重要组成部分,包括标准数据库建立、船底形变检测算法以及船舶三维吃水模型修正算法,并对算法进行了仿真测试。
最后给出了系统实船比对实验的结果,确定了系统的比对精度。结果表明,系统的比对精度为±0.048m,优于设计要求的0.1m,满足设计要求。采用的算法实用,有效。
China's water transport is very convenient, which leads the rapid development of inland water transport and a strong impetus to the development of national economy. With the increasing amount of inland water transport, inland vessels ultra-draft violations is also increasing, which make a serious thread on the inland waterways navigation safety and navigation efficiency, hindering the healthy development of the economy. Relevant law enforcement departments, however, still take the only way on board to make the detection of ship waterline to measurement the ship ultra-draft, which is both time-consuming and affects the normal navigation of the ship. At present, no effective techniques for rapid ship draft measurement.
Measurement of ship's3D draft can gain not only the ship's largest draft value, but also a lot of information of ship bottom model, which is the base of testing the deformation of ship's bottom and detecting the ship overload. According to the requirements of navigation ship's3D draft detection, we developed the navigation ship's3D draft dynamic measurement system based on single beam ultrasonic sensor array, ensure the premise of navigation efficiency, realize the parallel navigation ships3D draft dynamic test, and for the ultra-draft warning. At present the system has set up in the Yangtze three gorges shiplock and operates well.
The ultrasonic sensor array of measurement system is installed on the detection stent, using the detection trestle synchronous automatic lift subsystems putting down the trestle synchronously, smoothly to the measuring position underwater, avoid affecting the normal navigation. Ship's3D data acquisition and processing subsystem gains ship's3D draft data by sensor array, eliminate the abnormal data, make segmentation of the ship draft data and modification processing of3D draft data model. The draft data processed will be displayed in3D way on a computer, and every ship be measured will be calculated the deepest ship draft correspondingly. By a wireless network the result will be sent to the central management information system for the ultra-draft warning.
Use navigation ship's3D draft dynamic measurement system, we obtain the original ship's3D draft data model, which is with so much abnormal data in it and mix several ship's3D draft data together. For the original3D draft data model, use the point cloud data median filtering algorithm to eliminate the "similar random noise" and the least square method linear regression forecast method to eliminate "ship wake turbulence noise". For the preprocessed ship's3D draft data, this paper analyzes the features of hull in different parts of the draft data, and based on the inspection system dynamic measurement requirements, put forward the point cloud data segmentation algorithm based on the data features extraction to realize the3D draft data model dynamic segmentation, and use the actual measurement system data to verify the whole algorithm. At the same time, this paper introduces the dynamic3D display technology of ship draft data based on OpenGL, and uses the system measured data to verify the function.
In order to further enhance the system precision, the paper puts forward using the model matching technology realize ship bottom deformation testing and ship's3D draft revised. This paper introduces the important part of ship model matching technology, including standard database established, the deformation detection algorithm and3D model ship draft correction algorithm, and demonstrates the algorithm simulation test.
The system experiment result is given to determine the accuracy ratio. The results show that the precision of the system is±0.048meters, better than the requirement of design, which is0.1meters, meet the design requirements. The algorithm of the system is practical and effective.
检测船舶的三维吃水不仅能够获得船舶的最大吃水值,也可以获取船舶三维吃水数据模型等大量信息,是检测船舶超载以及检测船底形变的数据基础。根据通航船舶三维吃水的检测要求,研制了基于单波束超声波传感器阵列的通航船舶三维吃水动态检测系统,在保证通航效率的前提下,实现了对并行通航船舶三维吃水的快速离船式动态检测,并对超吃水船舶进行预警。目前系统已在长江三峡船闸投入使用,运行状态良好。
检测系统的超声波传感器阵列安装于检测支架上,采用检测门同步自动升降子系统将检测支架同步、平稳地放置于水下合适的测量位置,避免影响船舶正常通航。船舶三维吃水数据采集和处理子系统通过传感器阵列获得船舶三维吃水数据,进行异常数据剔除、多船吃水数据分割以及三维吃水数据模型修正处理,将数据处理后的吃水数据以三维形式显示于计算机上,并计算出每艘被测船舶对应的最大吃水,用无线网络发送至信息综合管理系统进行超吃水预警。
利用通航船舶三维吃水动态检测系统,获取船舶原始的三维吃水数据模型,其中存在较多的异常数据并且混合了多艘被测船舶的吃水数据。对于原始的三维吃水数据模型,采用点云数据中值滤波算法剔除“类随机噪声”,采用最小二乘法直线回归预测法剔除“船舶尾流噪声”。对于预处理后的船舶三维吃水数据,本文分析了船体不同部位的吃水数据特征,并结合系统的动态检测要求,提出了基于数据特提取的点云数据动态分割算法,对三维吃水数据模型进行动态分割,分离出并行通航船舶各自的三维吃水数据模型,并采用系统实际测量数据,对整体算法进行了验证。同时,本文介绍了基于OpenGL的船舶吃水数据动态三维显示技术,并采用系统实测数据进行了功能验证。
为了进一步提高系统的检测精度,本文提出采用船舶三维吃水模型匹配技术实现船舶底部形变检测以及船舶三维吃水模型的修正。本文详细介绍了船舶三维吃水模型匹配技术的重要组成部分,包括标准数据库建立、船底形变检测算法以及船舶三维吃水模型修正算法,并对算法进行了仿真测试。
最后给出了系统实船比对实验的结果,确定了系统的比对精度。结果表明,系统的比对精度为±0.048m,优于设计要求的0.1m,满足设计要求。采用的算法实用,有效。
China's water transport is very convenient, which leads the rapid development of inland water transport and a strong impetus to the development of national economy. With the increasing amount of inland water transport, inland vessels ultra-draft violations is also increasing, which make a serious thread on the inland waterways navigation safety and navigation efficiency, hindering the healthy development of the economy. Relevant law enforcement departments, however, still take the only way on board to make the detection of ship waterline to measurement the ship ultra-draft, which is both time-consuming and affects the normal navigation of the ship. At present, no effective techniques for rapid ship draft measurement.
Measurement of ship's3D draft can gain not only the ship's largest draft value, but also a lot of information of ship bottom model, which is the base of testing the deformation of ship's bottom and detecting the ship overload. According to the requirements of navigation ship's3D draft detection, we developed the navigation ship's3D draft dynamic measurement system based on single beam ultrasonic sensor array, ensure the premise of navigation efficiency, realize the parallel navigation ships3D draft dynamic test, and for the ultra-draft warning. At present the system has set up in the Yangtze three gorges shiplock and operates well.
The ultrasonic sensor array of measurement system is installed on the detection stent, using the detection trestle synchronous automatic lift subsystems putting down the trestle synchronously, smoothly to the measuring position underwater, avoid affecting the normal navigation. Ship's3D data acquisition and processing subsystem gains ship's3D draft data by sensor array, eliminate the abnormal data, make segmentation of the ship draft data and modification processing of3D draft data model. The draft data processed will be displayed in3D way on a computer, and every ship be measured will be calculated the deepest ship draft correspondingly. By a wireless network the result will be sent to the central management information system for the ultra-draft warning.
Use navigation ship's3D draft dynamic measurement system, we obtain the original ship's3D draft data model, which is with so much abnormal data in it and mix several ship's3D draft data together. For the original3D draft data model, use the point cloud data median filtering algorithm to eliminate the "similar random noise" and the least square method linear regression forecast method to eliminate "ship wake turbulence noise". For the preprocessed ship's3D draft data, this paper analyzes the features of hull in different parts of the draft data, and based on the inspection system dynamic measurement requirements, put forward the point cloud data segmentation algorithm based on the data features extraction to realize the3D draft data model dynamic segmentation, and use the actual measurement system data to verify the whole algorithm. At the same time, this paper introduces the dynamic3D display technology of ship draft data based on OpenGL, and uses the system measured data to verify the function.
In order to further enhance the system precision, the paper puts forward using the model matching technology realize ship bottom deformation testing and ship's3D draft revised. This paper introduces the important part of ship model matching technology, including standard database established, the deformation detection algorithm and3D model ship draft correction algorithm, and demonstrates the algorithm simulation test.
The system experiment result is given to determine the accuracy ratio. The results show that the precision of the system is±0.048meters, better than the requirement of design, which is0.1meters, meet the design requirements. The algorithm of the system is practical and effective.
引文
[1]http://baike.baidu.com/view/4185.htm
[2]http://www.chinanews.com/cj/2011/01-30/2821711.shtml.
[3]童飞,曾文.“超吃水”航行管理问题研究[J],中国水运.2008,7.
[4]周广程.图象处理技术在船舶吃水自动检测系统中的应用(硕士学位论文).南京理工大学.2006,12.
[5]郭方.基于视频的船舶吃水线检测方法的研究：(硕士学位论文).大连海事大学.2010,6.
[6]苏华.船舶吃水动态检测方法的研究(硕士学位论文).武汉理工大学.2008.5.
[7]王勇,颜昌平.船舶吃水现场检测技术及方法探讨[J].中国水运报.2008,(8)：4445.
[8]朱四印.船舶动态吃水实时检测与数据处理关键技术的研究[D].大连：大连海事大学,2011.
[9]钟辉.三峡通航船舶吃水远程监测与预警系统的研究[D].大连：大连海事大学,2011.
[10]Lirakis C B, Bongiovanni K, P.Automated Multibeam Data Cleaning and Target Detection[C]. Oceans 2000, MTS/IEEE Conference and Exhibition,2000,(l):719-723.
[11]Ulrich Lenk. Multibeam Data Processing:A Detailed Algorithm to Compute(Adaptive) Triangulations and Potential Indications of Data Quality[J].The Hydrographical Journal.2003,(107):3-10.
[12]李明叁,张杰,严怀志等.序统计滤波估计检测海洋测深异常数据[J].海洋测绘.2007,(27)：7-11.
[13]周达超.船舶吃水量信息采集处理系统的研究(硕士学位论文).大连海事大学.2010,6.
[14]刘春,陈华云，吴杭彬.激光三维遥感的数据处理与特征提取[M].北京：科学出版社.2009.
[15]王霞.点云曲面匹配的八叉树算法[J].计算机工程与应用.2009,45(32)：171-173,177.
[16]武殿梁等.基于遗传算法和最小二乘法的曲面匹配[J].航空学报.2002.23(3)：285-288.
[17]尹吉林,谢杰.一种改进的曲面配准方法的研究[J].研究与设计.2010,26(3)：1-3,5.
[18]王伯平，景大英.基于进化算法的空间面轮廓度误差的测量[J].仪器仪表学报.2005,26(8)：744-747.
[19]王东善,李明,卓军.一种自由曲面面轮廓度误差评定方法[J].机械设计与制造.2004,2(1)：64-65,66.
[20]孙坤凯.三峡船舶吃水量检测门同步升降系统的研究(硕士学位论文).大连海事大学.2010,6.
[21]三菱公司.Q系列CC-Link网络系统用户参考手册.2001,11.
[22]Peter Neumann. Communication in industrial automation—What is going on[J].Control Engineering Practice,2007,15(11):1332-1347.
[23]李伟.基于Modbus协议的工控节点设计与实现[J].计算机工程.2010,36(16)：266-228.
[24]庆兆洲科技公司.MH-C串口通信协议.2008,4.
[25]http://www.advantech.com.tw/epc/newsletter/atom_embedded_ipcs_20100205/index.html
[26]熊木地,朱四印,李禄,倪鹏.通航船舶吃水实时检测系统数据处理方法研究[J].2012-12,33(1)：173-180.
[27]鲍华,樊瑜波,饶长辉等.基于均值查找的快速中值滤波算法[J],四川大学学报(工程科学版).2011,43(2)：76-79.
[28]阮秋琦,阮余智等译Rafael C.Gonzalez,Richard E.Woods著.数字图像处理[M].北京：电子工业出版社.2007.
[29]G.Cao,Y.Zhao,R.Ni,et al. Forensic detection of median filtering in digital images, in IEEE Int. Conf. Multimedia and Expo (ICME), pp.89-94,2010
[30]F.Dong. A Novel Fast Algorithm for Image Median Filtering Based on Improved Sign Test Method, in IEEE Int. Conf. Logistics Systems and Intelligent Management, vol.1-3,pp.1240-1242,2010
[31]X.Deng, Y.Xiong, H,Peng. A New Kind of Weighted Median Filtering Algorithm Used For Processing, in Int. Symp. on Int.Science and Engineering, vol.2, pp.738-743,2008
[32]张莹,王耀南.基于局部加权偏最小二乘法的冷凝器污垢预测[J].仪器仪表学报,2010,31(2)：299-304.
[33]张起贵,张魁.基于最小二乘直线拟合的小目标检测[J].电子设计工程.2010,(07)：177
[34]田社平,丁国清.100数据拟合及其在测试数据处理中的应用[J].仪器仪表学报.2002,(z1)：3-4
[35]Delcroix M, Hikichi T, Miyoshi M. Dereverberation and denoising using multichannel linear prediction[J].IEEE Trans.on Audio, Speech, and Language Processing,2007,15(6):1791-1801.
[36]Jann N Yang, Hongwei Huang, Silian Lin. Sequential non-linear least-square estimation for damage identification of structures[J].International Journal of Non-Linear Mechanics,2006,41:124-140.
[37]刘少梅,杨鼎才.基于最小二乘法和主元分析的车牌倾斜校正方法[J].电子测量技术.2008,31(4)：49-51.
[38]Kozat, S.s.; Singer, A.C. Universal Switching Linear Least Squares Prediction, IEEE Trans. Signal Process. vol.56, no.1, pp.189-204,2008
[39]Papalexopoulos.A.D, Hesterberg.T.C. A Regression-based Approach To Short-term System Load Forecasting, IEEE Trans. Power Syst. Vol.5, no.4, pp.1535-1547,1990.
[40]Chi-Tsuen Yeh. Reduction to Least-Square Estimates in Multiple Fuzzy Regression Analysis, IEEE Tran. Fuzzy Syst., vol.17, no.4, pp.935-948,2009.
[41]刘进,武仲科,周明全,点云模型分割及应用技术综述[J].计算机科学.2011.38(4)：21-24,71
[42]李慧盈,李文辉,陈圣波.一种机载雷达点云数据的快速分类方法[J].吉林大学学报(地球科学版),2010,40(5)：1205-1210
[43]X.Wang, H. Wang, X.Ye et al. A Novel Segmentation Algorithm for Side-scan Sonar Imagery with Multi-object, in IEEE Int. conf. Robotics and Biomimetics, Dec 2007, pp.2110-2114.
[44]L.Sun. A Feature Extraction Method Suitable For Multi-channel Sensor Data, Int. Conf. Wavelet Analysis and Pattern Recognition,2007, vol.3, pp.1217-1220.
[45]Kumar.S, Ghosh.J, Crawford.M.M. Best-bases Feature Extraction Algorithms For Classification Of Hyperspectral Data, IEEE Trans. Geosci. Remote Sens. Vol.39, no.7, pp.1368-1379,Jul2001
[46]GChen, L.Ko, B.Kuo, S.Shih. A Tow-stage Feature Extraction For Hyperspectral Image Data Classification, in IEEE Int. Geoscience and Remote Sensing Symp.IGARSS'04 Proceedings,vol.2, pp.1212-1215,2004
[47]Mukherjee.K, Gupta.S, Ray.A et al. Symbolic Analysis of Sonar Data for Underwater Target Detection, J.Ocean.Eng. vol.36, no.2, pp.219-230
[48]李阳铭,孟庆虎.一种广泛适用的激光雷达数据特征提取方法[J].机器人.2010(6)：812-821.
[49]李必军,方志祥,任娟.从激光扫描数据中进行建筑物特征提取研究[J].武汉大学学报(信息科学版),2003,28(1)：65-69.
[50]吴世雄,王成勇.散乱噪声点云的数据分割[J].机械工程学报,2007,43(2)：230-233.
[51]http://baike.baidu.com/view/13769.htm
[52]胡伟,王弘.如何在VC++中用MFC进行OpenGL编程[J].计算机应用.2001,21(8)：87-89
[53]陈肇泰.测量结果的最佳估计及误差限[J].青岛教育学院学报.2001，14(2)：59-60.
[54]梁志国.数据采集系统误差限的测量不确定[J].计量技术.2002,(9)：45.-48.
[55]黄曙江.平均误差的置信度和极限误差[J].物理实验.1994,14(5)：221-222.
[56]蒋玉英,张元,葛宏义.微波检测粮仓水分技术的研究[J].计算机工程与应用.2010,46(29)：239-241.
[2]http://www.chinanews.com/cj/2011/01-30/2821711.shtml.
[3]童飞,曾文.“超吃水”航行管理问题研究[J],中国水运.2008,7.
[4]周广程.图象处理技术在船舶吃水自动检测系统中的应用(硕士学位论文).南京理工大学.2006,12.
[5]郭方.基于视频的船舶吃水线检测方法的研究：(硕士学位论文).大连海事大学.2010,6.
[6]苏华.船舶吃水动态检测方法的研究(硕士学位论文).武汉理工大学.2008.5.
[7]王勇,颜昌平.船舶吃水现场检测技术及方法探讨[J].中国水运报.2008,(8)：4445.
[8]朱四印.船舶动态吃水实时检测与数据处理关键技术的研究[D].大连：大连海事大学,2011.
[9]钟辉.三峡通航船舶吃水远程监测与预警系统的研究[D].大连：大连海事大学,2011.
[10]Lirakis C B, Bongiovanni K, P.Automated Multibeam Data Cleaning and Target Detection[C]. Oceans 2000, MTS/IEEE Conference and Exhibition,2000,(l):719-723.
[11]Ulrich Lenk. Multibeam Data Processing:A Detailed Algorithm to Compute(Adaptive) Triangulations and Potential Indications of Data Quality[J].The Hydrographical Journal.2003,(107):3-10.
[12]李明叁,张杰,严怀志等.序统计滤波估计检测海洋测深异常数据[J].海洋测绘.2007,(27)：7-11.
[13]周达超.船舶吃水量信息采集处理系统的研究(硕士学位论文).大连海事大学.2010,6.
[14]刘春,陈华云，吴杭彬.激光三维遥感的数据处理与特征提取[M].北京：科学出版社.2009.
[15]王霞.点云曲面匹配的八叉树算法[J].计算机工程与应用.2009,45(32)：171-173,177.
[16]武殿梁等.基于遗传算法和最小二乘法的曲面匹配[J].航空学报.2002.23(3)：285-288.
[17]尹吉林,谢杰.一种改进的曲面配准方法的研究[J].研究与设计.2010,26(3)：1-3,5.
[18]王伯平，景大英.基于进化算法的空间面轮廓度误差的测量[J].仪器仪表学报.2005,26(8)：744-747.
[19]王东善,李明,卓军.一种自由曲面面轮廓度误差评定方法[J].机械设计与制造.2004,2(1)：64-65,66.
[20]孙坤凯.三峡船舶吃水量检测门同步升降系统的研究(硕士学位论文).大连海事大学.2010,6.
[21]三菱公司.Q系列CC-Link网络系统用户参考手册.2001,11.
[22]Peter Neumann. Communication in industrial automation—What is going on[J].Control Engineering Practice,2007,15(11):1332-1347.
[23]李伟.基于Modbus协议的工控节点设计与实现[J].计算机工程.2010,36(16)：266-228.
[24]庆兆洲科技公司.MH-C串口通信协议.2008,4.
[25]http://www.advantech.com.tw/epc/newsletter/atom_embedded_ipcs_20100205/index.html
[26]熊木地,朱四印,李禄,倪鹏.通航船舶吃水实时检测系统数据处理方法研究[J].2012-12,33(1)：173-180.
[27]鲍华,樊瑜波,饶长辉等.基于均值查找的快速中值滤波算法[J],四川大学学报(工程科学版).2011,43(2)：76-79.
[28]阮秋琦,阮余智等译Rafael C.Gonzalez,Richard E.Woods著.数字图像处理[M].北京：电子工业出版社.2007.
[29]G.Cao,Y.Zhao,R.Ni,et al. Forensic detection of median filtering in digital images, in IEEE Int. Conf. Multimedia and Expo (ICME), pp.89-94,2010
[30]F.Dong. A Novel Fast Algorithm for Image Median Filtering Based on Improved Sign Test Method, in IEEE Int. Conf. Logistics Systems and Intelligent Management, vol.1-3,pp.1240-1242,2010
[31]X.Deng, Y.Xiong, H,Peng. A New Kind of Weighted Median Filtering Algorithm Used For Processing, in Int. Symp. on Int.Science and Engineering, vol.2, pp.738-743,2008
[32]张莹,王耀南.基于局部加权偏最小二乘法的冷凝器污垢预测[J].仪器仪表学报,2010,31(2)：299-304.
[33]张起贵,张魁.基于最小二乘直线拟合的小目标检测[J].电子设计工程.2010,(07)：177
[34]田社平,丁国清.100数据拟合及其在测试数据处理中的应用[J].仪器仪表学报.2002,(z1)：3-4
[35]Delcroix M, Hikichi T, Miyoshi M. Dereverberation and denoising using multichannel linear prediction[J].IEEE Trans.on Audio, Speech, and Language Processing,2007,15(6):1791-1801.
[36]Jann N Yang, Hongwei Huang, Silian Lin. Sequential non-linear least-square estimation for damage identification of structures[J].International Journal of Non-Linear Mechanics,2006,41:124-140.
[37]刘少梅,杨鼎才.基于最小二乘法和主元分析的车牌倾斜校正方法[J].电子测量技术.2008,31(4)：49-51.
[38]Kozat, S.s.; Singer, A.C. Universal Switching Linear Least Squares Prediction, IEEE Trans. Signal Process. vol.56, no.1, pp.189-204,2008
[39]Papalexopoulos.A.D, Hesterberg.T.C. A Regression-based Approach To Short-term System Load Forecasting, IEEE Trans. Power Syst. Vol.5, no.4, pp.1535-1547,1990.
[40]Chi-Tsuen Yeh. Reduction to Least-Square Estimates in Multiple Fuzzy Regression Analysis, IEEE Tran. Fuzzy Syst., vol.17, no.4, pp.935-948,2009.
[41]刘进,武仲科,周明全,点云模型分割及应用技术综述[J].计算机科学.2011.38(4)：21-24,71
[42]李慧盈,李文辉,陈圣波.一种机载雷达点云数据的快速分类方法[J].吉林大学学报(地球科学版),2010,40(5)：1205-1210
[43]X.Wang, H. Wang, X.Ye et al. A Novel Segmentation Algorithm for Side-scan Sonar Imagery with Multi-object, in IEEE Int. conf. Robotics and Biomimetics, Dec 2007, pp.2110-2114.
[44]L.Sun. A Feature Extraction Method Suitable For Multi-channel Sensor Data, Int. Conf. Wavelet Analysis and Pattern Recognition,2007, vol.3, pp.1217-1220.
[45]Kumar.S, Ghosh.J, Crawford.M.M. Best-bases Feature Extraction Algorithms For Classification Of Hyperspectral Data, IEEE Trans. Geosci. Remote Sens. Vol.39, no.7, pp.1368-1379,Jul2001
[46]GChen, L.Ko, B.Kuo, S.Shih. A Tow-stage Feature Extraction For Hyperspectral Image Data Classification, in IEEE Int. Geoscience and Remote Sensing Symp.IGARSS'04 Proceedings,vol.2, pp.1212-1215,2004
[47]Mukherjee.K, Gupta.S, Ray.A et al. Symbolic Analysis of Sonar Data for Underwater Target Detection, J.Ocean.Eng. vol.36, no.2, pp.219-230
[48]李阳铭,孟庆虎.一种广泛适用的激光雷达数据特征提取方法[J].机器人.2010(6)：812-821.
[49]李必军,方志祥,任娟.从激光扫描数据中进行建筑物特征提取研究[J].武汉大学学报(信息科学版),2003,28(1)：65-69.
[50]吴世雄,王成勇.散乱噪声点云的数据分割[J].机械工程学报,2007,43(2)：230-233.
[51]http://baike.baidu.com/view/13769.htm
[52]胡伟,王弘.如何在VC++中用MFC进行OpenGL编程[J].计算机应用.2001,21(8)：87-89
[53]陈肇泰.测量结果的最佳估计及误差限[J].青岛教育学院学报.2001，14(2)：59-60.
[54]梁志国.数据采集系统误差限的测量不确定[J].计量技术.2002,(9)：45.-48.
[55]黄曙江.平均误差的置信度和极限误差[J].物理实验.1994,14(5)：221-222.
[56]蒋玉英,张元,葛宏义.微波检测粮仓水分技术的研究[J].计算机工程与应用.2010,46(29)：239-241.