运用三维声纳图像的水底全景拼接与绘制Submarine global mosaicing and plotting based on 3D sonar imaging
李志华,李秋峦
摘要(Abstract)：
针对海量声纳图像多帧配准引起的误差累积问题,该文提出了一种基于三维声纳图像的水底全景拼接与绘制方法。首先通过相邻两帧图像最近邻点迭代,并结合GPS、姿态仪信息和X84控制点剔除方法,减少非重叠区域控制点对的存在以提高配准精度,实现相邻帧图像精确配准。然后采用四元组参数对所有精确配准矩阵进行拟牛顿最优化处理,获取所有声纳帧之间最佳配准关系,减少多帧配准引起的累积误差,实现三维图像高精度全局配准。湖试与海试实验结果表明,该方法有效地减少了海量声纳图像多帧配准引起的累积误差,实现了三维声纳图像水底全景高精度拼接与绘制,达到了预期目标。
关键词(KeyWords)： 三维声纳;图像配准;图像拼接;图像绘制
基金项目(Foundation): 国家自然科学基金项目(61001170)
作者(Author): 李志华,李秋峦
DOI: 10.16251/j.cnki.1009-2307.2015.06.0030
参考文献(References)：
- [1]李海森,陈宝伟,么彬,等.多子阵高分辨海底地形探测算法及其FPGA和DSP阵列实现[J].仪器仪表学报,2010,31(2):281-285.
- [2]梅赛,高金耀,杨春国,等.冲绳海槽南部多波束海底地形虚拟视景仿真初探[J].海洋测绘,2010,30(2):19-23.
- [3]Kaiho I.A Dream for Realization of High-speed,3D Volume Image by Lens-focused Method[J].International Congress Series,2004(1274):38-46.
- [4]Hansen R K,Castellani U,Murino V,et al.Mosaicing of3DSonar Data Sets-Techniques and Applications[C]//Proc.IEEE OCEANS MTS.Washington,DC,USA:IEEE Computer Society,Piscataway,USA,2005:2326-2333.
- [5]Yu S C,Kim T W,Marani G,et al.Real-time 3DSonar Image Recognition for Underwater Vehicles[C]//Proc.Symposium on Underwater Technology and Workshop on Scientific Use of Submarine Cables and Related Technologies.2007:142-146.
- [6]Repetto S,Palmese M,Trucco A.High-resolution 3-D Imaging by a Sparse Array:Array Optimization and Image Simulation[C]//Proc.Oceans.2005:763-768.
- [7]Baranoski E J.Real-time Array Signal Processors for Embedded Applications[C]//Proc.the Thirty-Third Asilomar Conference on Signals,Systems,and Computers.1999:1381-1385.
- [8]Murino V,Trucco A.Three-dimensional Image Generation and Processing in Underwater Acoustic Vision[J].Proceedings of IEEE.2000,88(12):1903-1948.
- [9]Ikeuchi K,Wheeler M D.Iterative Estimation of Rotation and Translation Using Quaternion[R].Technical Report CMU-SC-95-215,Carnegie Mellon University,1995.
- [10]Zhang Z.Iterative Point Matching for Free-form Curves and Surfaces[J].International Journal of Computer Vision,1994,13(2):119-152.
- [11]Codaoctopus.High Resolution 3-dimensional Images in Real Time Unique Sonar Technology[EB/OL].http://www.codaoctopus.com/3d_ac_im/index.asp.
- [12]Cignoni P,Montani C,Scopigno R,et al.The Marching Intersections Algorithm for Merging Range Images[R].Technical Report B4-61-00,I.E.I.-C.N.R.,Pisa,Italy,2000.