基于侧扫声呐图像的水下3维地形环境重建

Underwater 3D Terrain Environment Reconstruction Based on Side-scan Sonar Images

  • 摘要: 为了能够360°展现海底地形环境和海底目标的3维立体空间形态,设计了基于BFGS(Broyden-Fletcher-Goldfarb-Shanno)法和交替方向乘子法的明暗恢复形状(SFS)3维重建方法。通过该方法和预处理共轭梯度迭代优化算法对2维侧扫声呐图像进行迭代,处理相关约束条件,确保最终3维重建结果的一致性与平滑性。得到3维重建的水下地形图后,建立了3维重建算法指标,同时进行了一系列实验验证,确保所提出的SFS三维重建方法不仅在理论上切实可行,而且实践效果显著,满足水下3维地形重建的需求。

     

    Abstract: In order to reveal the 3D spatial morphology of the seabed terrain environment and seabed targets in 360°, a 3D reconstruction method is designed based on BFGS(Broyden-Fletcher-Goldfarb-Shanno) algorithm and ADMM(alternating direction method of multipliers) for SFS(shape from shading). The BFGS-ADMM and PCG(preconditioned conjugate gradient) iterative optimization algorithms are used to iterate the 2D side-scan sonar images and process the relevant constraints to ensure the consistency and smoothness of the final 3D reconstruction results. After obtaining the 3D reconstructed underwater topography, the 3D reconstruction algorithm index is established, and a series of experimental verification is carried out at the same time to ensure that the proposed SFS 3D reconstruction method is not only theoretically practical and feasible, but also practically effective, and meets the needs of underwater 3D topography reconstruction.

     

/

返回文章
返回