三维成像声纳成像算法仿真研究
|
摘要
随着人们对水下探测需求的增长,传统的二维成像声呐已渐渐不能满足人们的要求,三维成像声呐由于能提供水下目标的三维信息,得到的图像更直观,受到了越来越多关注。而目前常用的水下三维成像方法,需要经过多次多角度重复测量,或者需要较复杂的面阵或声透镜。为此,本文研究一种简化的面阵-L型基阵对水下目标进行三维成像的方法。
本文改进了一种基于常规的波束形成技术的三维成像方法。将两条线阵的多波束接收数据,按照一定的顺序一一对应相乘,生成一组新数据,再通过对该数据的处理,提取出目标的三维信息,就可以得到目标的三维图像。同时,还从两个不同角度得到了目标的二维图像,从这两幅图像中,也可以提取出目标的三维信息。
多子阵波束Root-MUSIC(MSB-RMU)算法较常规高分辨方位估计方法需要的信噪比条件低,又能使空间分辨力不受阵列孔径的限制。在本文中,利用L型基阵这种特殊的阵列形式,将一维MSB-RMU算法扩展到二维,用于三维成像,还对采用的参数配对方法做了进一步分析,结合成像仿真结果,采用了分组排序的参数配对方法。另外,还采用了一种结合了MSB-RMU算法与二维MUSIC算法的参数自动配对的算法。
最后,通过建立的仿真成像场景,对上述几种三维成像方法的成像效果进行了对比分析。采用二维MSB-RMU算法的成像效果要明显好于前两种方法。其中,对强度不相等的信源来说,两种采用高分辨技术的算法的成像效果相似,不过,参数自动配对的DOA估计方法对等强度信源同样有效,而二维MSB-RMU算法则会失效。
As the conventional 2-D imaging sonar can not fully meet the demands for underwater detection anymore, the sonar system that can provide real-time 3-D images underwater become more and more important. But the real-time 3-D imaging sonar usually depends on 2-D planar array or acoustic lens technology, which are much more complex and expensive than linear array. In this paper, some new 3-D imaging methods that depend on an L shaped array are proposed.
A new method that makes use of conventional beam-forming technology is proposed. Its basic principle is multiplying the beam-forming data received from one of the linear array and the beam-forming data received from another linear array together to forming a group of new data, and then forming the 3-D imaging of the object region based on this new data. 2-D images can also be drawn at the same time. With the help of 2-D images, the 3-D imaging of the object region can be more complete.
MSB-RMU algorithm is a high resolution DOA estimation algorithm with less SNR requirements than conventional ones. In this paper, 1-D MSB-RMU algorithm develops into 2-D MSB-RMU algorithm by using L shaped array. It can be used for 3-D imaging after parameter matching. Two parameter matching methods are applied. They both depend on the estimation of the strength of target and have some differences in the procedure of matching. Then a new method that combines 1-D MSB-RMU algorithm with 2-D MUSIC algorithm is mentioned too. There is no need to use parameter matching method in this new algorithm. And it can be used for equal-power targets estimation.
At last, a simulated imaging model is built and the methods mentioned before are applied. The results of algorithms that use high resolution DOA estimation technology turn out to be much better than the others. The two high resolution algorithms have the same performance to the targets of different strength. But 2-D MSB-RMU algorithm will fail if the targets are have same strength. And this will not happen to the other algorithm.
本文改进了一种基于常规的波束形成技术的三维成像方法。将两条线阵的多波束接收数据,按照一定的顺序一一对应相乘,生成一组新数据,再通过对该数据的处理,提取出目标的三维信息,就可以得到目标的三维图像。同时,还从两个不同角度得到了目标的二维图像,从这两幅图像中,也可以提取出目标的三维信息。
多子阵波束Root-MUSIC(MSB-RMU)算法较常规高分辨方位估计方法需要的信噪比条件低,又能使空间分辨力不受阵列孔径的限制。在本文中,利用L型基阵这种特殊的阵列形式,将一维MSB-RMU算法扩展到二维,用于三维成像,还对采用的参数配对方法做了进一步分析,结合成像仿真结果,采用了分组排序的参数配对方法。另外,还采用了一种结合了MSB-RMU算法与二维MUSIC算法的参数自动配对的算法。
最后,通过建立的仿真成像场景,对上述几种三维成像方法的成像效果进行了对比分析。采用二维MSB-RMU算法的成像效果要明显好于前两种方法。其中,对强度不相等的信源来说,两种采用高分辨技术的算法的成像效果相似,不过,参数自动配对的DOA估计方法对等强度信源同样有效,而二维MSB-RMU算法则会失效。
As the conventional 2-D imaging sonar can not fully meet the demands for underwater detection anymore, the sonar system that can provide real-time 3-D images underwater become more and more important. But the real-time 3-D imaging sonar usually depends on 2-D planar array or acoustic lens technology, which are much more complex and expensive than linear array. In this paper, some new 3-D imaging methods that depend on an L shaped array are proposed.
A new method that makes use of conventional beam-forming technology is proposed. Its basic principle is multiplying the beam-forming data received from one of the linear array and the beam-forming data received from another linear array together to forming a group of new data, and then forming the 3-D imaging of the object region based on this new data. 2-D images can also be drawn at the same time. With the help of 2-D images, the 3-D imaging of the object region can be more complete.
MSB-RMU algorithm is a high resolution DOA estimation algorithm with less SNR requirements than conventional ones. In this paper, 1-D MSB-RMU algorithm develops into 2-D MSB-RMU algorithm by using L shaped array. It can be used for 3-D imaging after parameter matching. Two parameter matching methods are applied. They both depend on the estimation of the strength of target and have some differences in the procedure of matching. Then a new method that combines 1-D MSB-RMU algorithm with 2-D MUSIC algorithm is mentioned too. There is no need to use parameter matching method in this new algorithm. And it can be used for equal-power targets estimation.
At last, a simulated imaging model is built and the methods mentioned before are applied. The results of algorithms that use high resolution DOA estimation technology turn out to be much better than the others. The two high resolution algorithms have the same performance to the targets of different strength. But 2-D MSB-RMU algorithm will fail if the targets are have same strength. And this will not happen to the other algorithm.
引文
[1]张小平.高分辨率多波束成像声呐关键技术研究.哈尔滨工程大学博士学位论文, 2005:1-2页
[2]李婷婷.三维前视声呐的高分辨方位估计研究.哈尔滨工程大学硕士学位论文, 2009:1-2页
[3] Anthony Davis,Angus Lugsdin.High speed underwater inspection for port and harbour security using Coda Echoscope 3D sonar.Oceans 2005, 2005,3: 2006-2011P
[4] A.Lorenson and D.Kraus.3D-Sonar image formation and shape recognition techniques.Oceans 2009-Europe, 2009:6P
[5]王彪,田杰,张春华.水声图象目标探测.探测与控制学报, 2004,26(4):34-37页
[6] William H.Key.Side scan sonar technology.Oceans 2000 MTS/IEEE Conference and Exhibition.Conference Proceedings,2000,2:1029 - 1033P
[7]张洁.基于声呐的水下机器人同时定位于地图构建技术研究.中国海洋大学硕士学位论文, 2008:21-22,30页
[8]张铁栋.前视声呐的后置图像处理算法研究.哈尔滨工程大学硕士学位论文,2004:2-3页
[9]张铁栋,万磊,马悦.一种单波束前视声呐成像方法.声学与电子工程, 2008,3:14-18页
[10]袁连喜.水下智能机器人声视觉成像可视化技术研究.哈尔滨工程大学硕士学位论文, 2002:6-7页
[11]唐利娜.水下三维声成像及声呐图像处理技术研究.哈尔滨工程大学硕士学位论文, 2009:2-4页
[12] Son-Cheol Yu,Tae-Won Kim,Giacomo Marani,Song K.Choi.Real-time 3D sonar image recognition for underwater vehicles.International Workshop on Scientific Use of Submarine Cables and RelatedTechnologies,2007:42-46P
[13]刘东宇.声透镜波束形成和水声图像处理技术研究.哈尔滨工程大学硕士学位论文. 2009:1-2页
[14] Edward O . Belcher , Dana C . Lynn , Hien Q . Dinh, Thomas J.Laughlin.Beamforming and imaging with acoustic lenses in small, high-frequency sonars.Oceans '99.MTS/IEEE.Riding the Crest into the 21st Century.Conference and Exhibition. Conference Proceedings,1999,3:1495-1499P
[15] CodaOctopus Group Inc.ECHOSCOPE II Product Brief.2009:1-2P
[16] Edward O.Belcher,CWO2 Jeffery R.Barone,Dennis G.Gallagher,Ronald E.Honaker.Acoustic lens camera and underwater display combine to provide efficient and effective hull and berth inspections. MTS Proceedings of Oceans’03 Coference,2003,3:1361-1367P
[17] Sound Metrics Corp.DISION Product Brief.2007:1-2P
[18]朱维庆,刘晓东,张东升,张方生,秦高林,廖诤,方常乐.高分辨率测深侧扫声呐.海洋技术. 2005,24(4):29-35页
[19]马爱民,郭伟民,洪星.高频声呐回波影像仿真技术.系统仿真学报. 2006,18(12):3342-3345页
[20]丁迎迎.海底物体回波模拟与图像生成技术研究.西北工业大学硕士学位论文. 2006:22,30-32页
[21]杨蕾.时延_相移波束形成技术研究.哈尔滨工程大学硕士学位论文,2004:14,20页
[22]田坦,刘国枝,孙大军.声呐技术.哈尔滨工程大学出版社,2000: 77,38页
[23]梁亦慧.数字波束形成算法仿真实现.应用光学,2003,24(3):43-45页
[24]王晓峰.高频二维成像声呐硬件系统的设计与实现.哈尔滨工程大学硕士学位论文,2004:38页
[25]田坦.水下定位与导航技术.国防工业出版社,2007:133页
[26] A. D. Whalen著,刘其培,迟惠生译.噪声中的信号检测(第二版).科学出版社,1977:64-74页
[27]王一童.前视扫描声呐的成像与目标特征提取.中国海洋大学硕士学位论文,2009:15页
[28]吴英姿.多波束测深系统地形跟踪与数据处理技术研究.哈尔滨工程大学博士学位论文,2001:36,68页
[29] Yuefeng LU,Masaki OSHIMA.On the 3-D reconstruction of seabed using multiple sidescan sonar images.IAPR workshop on machine vision applications,2002:584-587P
[30]王琪,刘雁春,暴景阳.关于多波束与侧扫声呐回波检测方法的探讨.海洋技术,2003,22(3):70-73页
[31]冯西安.水下目标高分辨方位估计技术研究.西北工业大学博士学位论文,2004:24页
[32]王永良,陈辉,彭应宁,万群.空间谱估计理论与算法.清华大学出版社,2004:19页
[33]包志强.快速稳健的参数估计及波束形成技术研究.西安电子科技大学博士学位论文,2006:24-26页
[34] Bhaskar D.Rao,K.V.S.Hari.Performance analysis of root-music.IEEE Transactions on Acoustics,Speech and Signal Processing,Dec.1989, 37(12):1939-1949P
[35]冯杰稳.健波束形成与高分辨方位估计技术研究.西北工业大学博士学位论文,2006:50页
[36] Michael D.Zoltowski,Gregory M.Kautz,Seth D.Silverstein. Beamspace root-music.IEEE Transactions on Signal Processing,Jan.1993,41(1):344-364P
[37] Luren Yang,Torfinn Taxt.Multibeam sonar bottom detection using multiple subarrays.Oceans '97.MTS/IEEE. Conference Proceedings,1997,2:932-938P
[38]黎子盛.多子阵波束域海底地形地貌高分辨探测技术.哈尔滨工程大学硕士学位论文,2007:56-60页
[39] Yingbo Hua,Tapan K.Sarkar,Donald D.Weiner.An L-shaped arrayfor estimating 2-D directions of wave arrival.IEEE Transactions on Antennas and Propagation,Feb.1991,39(2):143-146P
[40]李启虎.声呐信号处理引论(第二版).海洋出版社,2000:161页
[41]王朝英.测深侧扫声呐仿真研究.哈尔滨工程大学硕士学位论文,2007:16,17页
[42]姜永珉,冯海泓,郝新亚,惠俊英.水中目标二维亮点分布研究.声学学报,1997,22(1):79-86页
[43]杨海.分布式目标建模及其高分辨方位估计.西北工业大学硕士学位论文,2001:30-31页
[44]范军.水下复杂目标回声特性研究.上海交通大学博士学位论文,2001:16-18页
[45]胡涛.水下目标可视化仿真技术研究.哈尔滨工程大学硕士学位论文,2007:14-17页
[46]汤渭霖.声呐目标回波的亮点模型.声学学报,1994,19(2):93-100页
[47]赵春晖,胡涛,付正威.水下目标声散射特性仿真.应用科技,2007,34(6):1-4页
[48]王新晓,黄建国,马国强,胡方,张永峰.基于多亮点模型的目标回波信号在水声对抗仿真系统中的应用.探测与控制学报,2005,24(3):55-58页
[49] P.H.Kraeutner,J.S.Bird,Bill Charbonneau,Dan Bishop,Fred Hegg.Multi-Angle swath bathymetry sidescan quantitative performance analysis.Oceans Conference Record (IEEE),2002,4:2253-2263P
[50] Vittotio Murino,Andrea Trucco.Three dimensional image generation and processing in underwater acoustic vision.Proceedings of the IEEE,Dec.2000,88(12):1903-1948P
[51]黎子盛,张建华,周天,李海森.基于L型阵列的二维CAATI算法高分辨DOA估计.海洋测绘,2007,27(1):27-29页
[52] A. D. Waite.实用声呐工程(第三版).电子工业出版社,2004:31页
[2]李婷婷.三维前视声呐的高分辨方位估计研究.哈尔滨工程大学硕士学位论文, 2009:1-2页
[3] Anthony Davis,Angus Lugsdin.High speed underwater inspection for port and harbour security using Coda Echoscope 3D sonar.Oceans 2005, 2005,3: 2006-2011P
[4] A.Lorenson and D.Kraus.3D-Sonar image formation and shape recognition techniques.Oceans 2009-Europe, 2009:6P
[5]王彪,田杰,张春华.水声图象目标探测.探测与控制学报, 2004,26(4):34-37页
[6] William H.Key.Side scan sonar technology.Oceans 2000 MTS/IEEE Conference and Exhibition.Conference Proceedings,2000,2:1029 - 1033P
[7]张洁.基于声呐的水下机器人同时定位于地图构建技术研究.中国海洋大学硕士学位论文, 2008:21-22,30页
[8]张铁栋.前视声呐的后置图像处理算法研究.哈尔滨工程大学硕士学位论文,2004:2-3页
[9]张铁栋,万磊,马悦.一种单波束前视声呐成像方法.声学与电子工程, 2008,3:14-18页
[10]袁连喜.水下智能机器人声视觉成像可视化技术研究.哈尔滨工程大学硕士学位论文, 2002:6-7页
[11]唐利娜.水下三维声成像及声呐图像处理技术研究.哈尔滨工程大学硕士学位论文, 2009:2-4页
[12] Son-Cheol Yu,Tae-Won Kim,Giacomo Marani,Song K.Choi.Real-time 3D sonar image recognition for underwater vehicles.International Workshop on Scientific Use of Submarine Cables and RelatedTechnologies,2007:42-46P
[13]刘东宇.声透镜波束形成和水声图像处理技术研究.哈尔滨工程大学硕士学位论文. 2009:1-2页
[14] Edward O . Belcher , Dana C . Lynn , Hien Q . Dinh, Thomas J.Laughlin.Beamforming and imaging with acoustic lenses in small, high-frequency sonars.Oceans '99.MTS/IEEE.Riding the Crest into the 21st Century.Conference and Exhibition. Conference Proceedings,1999,3:1495-1499P
[15] CodaOctopus Group Inc.ECHOSCOPE II Product Brief.2009:1-2P
[16] Edward O.Belcher,CWO2 Jeffery R.Barone,Dennis G.Gallagher,Ronald E.Honaker.Acoustic lens camera and underwater display combine to provide efficient and effective hull and berth inspections. MTS Proceedings of Oceans’03 Coference,2003,3:1361-1367P
[17] Sound Metrics Corp.DISION Product Brief.2007:1-2P
[18]朱维庆,刘晓东,张东升,张方生,秦高林,廖诤,方常乐.高分辨率测深侧扫声呐.海洋技术. 2005,24(4):29-35页
[19]马爱民,郭伟民,洪星.高频声呐回波影像仿真技术.系统仿真学报. 2006,18(12):3342-3345页
[20]丁迎迎.海底物体回波模拟与图像生成技术研究.西北工业大学硕士学位论文. 2006:22,30-32页
[21]杨蕾.时延_相移波束形成技术研究.哈尔滨工程大学硕士学位论文,2004:14,20页
[22]田坦,刘国枝,孙大军.声呐技术.哈尔滨工程大学出版社,2000: 77,38页
[23]梁亦慧.数字波束形成算法仿真实现.应用光学,2003,24(3):43-45页
[24]王晓峰.高频二维成像声呐硬件系统的设计与实现.哈尔滨工程大学硕士学位论文,2004:38页
[25]田坦.水下定位与导航技术.国防工业出版社,2007:133页
[26] A. D. Whalen著,刘其培,迟惠生译.噪声中的信号检测(第二版).科学出版社,1977:64-74页
[27]王一童.前视扫描声呐的成像与目标特征提取.中国海洋大学硕士学位论文,2009:15页
[28]吴英姿.多波束测深系统地形跟踪与数据处理技术研究.哈尔滨工程大学博士学位论文,2001:36,68页
[29] Yuefeng LU,Masaki OSHIMA.On the 3-D reconstruction of seabed using multiple sidescan sonar images.IAPR workshop on machine vision applications,2002:584-587P
[30]王琪,刘雁春,暴景阳.关于多波束与侧扫声呐回波检测方法的探讨.海洋技术,2003,22(3):70-73页
[31]冯西安.水下目标高分辨方位估计技术研究.西北工业大学博士学位论文,2004:24页
[32]王永良,陈辉,彭应宁,万群.空间谱估计理论与算法.清华大学出版社,2004:19页
[33]包志强.快速稳健的参数估计及波束形成技术研究.西安电子科技大学博士学位论文,2006:24-26页
[34] Bhaskar D.Rao,K.V.S.Hari.Performance analysis of root-music.IEEE Transactions on Acoustics,Speech and Signal Processing,Dec.1989, 37(12):1939-1949P
[35]冯杰稳.健波束形成与高分辨方位估计技术研究.西北工业大学博士学位论文,2006:50页
[36] Michael D.Zoltowski,Gregory M.Kautz,Seth D.Silverstein. Beamspace root-music.IEEE Transactions on Signal Processing,Jan.1993,41(1):344-364P
[37] Luren Yang,Torfinn Taxt.Multibeam sonar bottom detection using multiple subarrays.Oceans '97.MTS/IEEE. Conference Proceedings,1997,2:932-938P
[38]黎子盛.多子阵波束域海底地形地貌高分辨探测技术.哈尔滨工程大学硕士学位论文,2007:56-60页
[39] Yingbo Hua,Tapan K.Sarkar,Donald D.Weiner.An L-shaped arrayfor estimating 2-D directions of wave arrival.IEEE Transactions on Antennas and Propagation,Feb.1991,39(2):143-146P
[40]李启虎.声呐信号处理引论(第二版).海洋出版社,2000:161页
[41]王朝英.测深侧扫声呐仿真研究.哈尔滨工程大学硕士学位论文,2007:16,17页
[42]姜永珉,冯海泓,郝新亚,惠俊英.水中目标二维亮点分布研究.声学学报,1997,22(1):79-86页
[43]杨海.分布式目标建模及其高分辨方位估计.西北工业大学硕士学位论文,2001:30-31页
[44]范军.水下复杂目标回声特性研究.上海交通大学博士学位论文,2001:16-18页
[45]胡涛.水下目标可视化仿真技术研究.哈尔滨工程大学硕士学位论文,2007:14-17页
[46]汤渭霖.声呐目标回波的亮点模型.声学学报,1994,19(2):93-100页
[47]赵春晖,胡涛,付正威.水下目标声散射特性仿真.应用科技,2007,34(6):1-4页
[48]王新晓,黄建国,马国强,胡方,张永峰.基于多亮点模型的目标回波信号在水声对抗仿真系统中的应用.探测与控制学报,2005,24(3):55-58页
[49] P.H.Kraeutner,J.S.Bird,Bill Charbonneau,Dan Bishop,Fred Hegg.Multi-Angle swath bathymetry sidescan quantitative performance analysis.Oceans Conference Record (IEEE),2002,4:2253-2263P
[50] Vittotio Murino,Andrea Trucco.Three dimensional image generation and processing in underwater acoustic vision.Proceedings of the IEEE,Dec.2000,88(12):1903-1948P
[51]黎子盛,张建华,周天,李海森.基于L型阵列的二维CAATI算法高分辨DOA估计.海洋测绘,2007,27(1):27-29页
[52] A. D. Waite.实用声呐工程(第三版).电子工业出版社,2004:31页