节点文献
基于双探头多波束内部重叠区约束的横摇安装残差探测与校正
Detection and calibration of roll installation residuals based on constraints of dual-head multibeam overlapping regions
【摘要】 针对双探头多波束测深系统换能器安装校正后仍存在角度残差的问题,本研究提出一种基于双探头多波束内部重叠区约束换能器横摇安装残差探测与校正方法。首先分析测量过程中横摇安装残差的存在对海底地形的影响,利用相邻测线局部平坦地形数据为约束,采用随机抽样一致(RANSAC)算法进行平面拟合,探测并获取两平面夹角,实现一侧探头横摇安装残差校正;然后,基于双探头多波束同一测线两探头内部重叠区域为约束,利用RANSAC算法探测和获取另一侧探头横摇安装残差;最后,完成双探头多波束横摇安装残差的精确校正。实验结果表明:本算法可有效探测横摇安装残差,改正后相邻测线重叠区域水深不符值小于0.5 m的点云数占99%以上,其中测线间均值为0.043 m,标准差为0.003 m,优于Patch test的0.099和0.008 m;测线内两探头重叠区水深不符值的均值为0.006 m,标准差为0.006 m,优于Patch test的0.009和0.021 m,具有良好一致性。
【Abstract】 To solve the problem that there is still an angular deviation after the installation and calibration of the transducer in the dual-head multibeam sounding system, this study proposed a method for the detection and calibration of roll installation residuals of transducers based on the constraints of the dual-head multibeam internal overlapping regions. Firstly, the influence of the existence of roll installation residuals on submarine topography during the measurement process was analyzed, and using the local flat terrain data of the adjacent survey line as the constraint and the random sample consensus(RANSAC) algorithm for planar fitting, the angle between the two planes was detected and obtained, thus realizing the calibration of the roll installation residuals on one side of the head. Then, based on the constraint of the internal overlapping region of the two heads of the same survey line with multiple beams and two heads, the RANSAC algorithm was used to detect and obtain the mounting residuals of the other side of the head. Finally, the accurate calibration of the dual-head multibeam roll installation residuals was completed. The experimental results show that the algorithm can effectively detect the roll installation residuals. The number of point clouds with water depth discrepancies less than 0.5 m in the overlapping areas of adjacent survey lines accounts for more than 99%. The mean value between survey lines is 0.043 m and the standard deviation is 0.003 m, which are better than that of the patch test of 0.099 m and 0.008 m. The mean value and standard deviation of the water depth discrepancy in the overlap area of the two heads in the survey line are both 0.006 m, which are better than the 0.009 m and 0.021 m of the patch test, indicating that the algorithm has good consistency.
【Key words】 dual-head multibeam; RANSAC algorithm; planar fitting; roll installation residuals; auto-calibration;
- 【文献出处】 山东科技大学学报(自然科学版) ,Journal of Shandong University of Science and Technology(Natural Science) , 编辑部邮箱 ,2025年01期
- 【分类号】P229
- 【下载频次】36