针对直接使用大数据量的离散深度数据实现海底地形可视化时,基于不规则三角网的海底地形建模速度慢、稳健性低的问题,提出一种联合矢积测试的子网合并双向缝合算法,设计了一种基于二叉树的动态分块合并机制。该算法能够避免相交三角形或线形三角形的产生并使缝合后新生成的初始三角形形状趋于最优,提高了子网合并速度与整个构网程序的稳健性,实现了基于快速Delaunay构网的海底地形建模和三维可视化。实验结果表明,该算法能够明显提高子网合并速度与程序稳健性,适用于大数据量离散测深数据的海底地形绘制。
Aiming at the problem of low efficiency and poor robustness of seafloor terrain modeling based on triangula- ted irregular network while using discrete bathymetric data to visualize seafloor terrain directly, a bidirectional sew algorithm combining cross product test is proposed, and a dynamic segmentation and combination mechanism based on binary tree is designed. The algorithm avoids triangles with intersection or line shape and obtains triangles with good shape while two subnets combine, enhances the efficiency of subnet combining and the robustness of the whole procedure, and achieves seafloor terrain modeling and visualization based on fast Delaunay triangulation. Experiment results show that the algorithm obviously enhances the efficiency and robustness of subnet combining, and it can be applied to seafloor terrain rendering from large volumes of discrete bathymetric data.