基于拟准检定法的多波束测深粗差探测

Outlier Detection of Multibeam Echosounding Based on Quasi-accurate Detection of Outlier

针对最小二乘趋势面滤波法在多波束测深粗差探测中存在的不足,本文提出将拟准检定法应用于多波束粗差探测中,进而将拟准检定法探测粗差与最小二乘趋势面拟合有机结合起来。通过构造不同阶次的趋势面,并对实测数据进行分析,结果表明:拟准检定法剔除粗差后海底地形表面较为平滑,微小地形得以保存,跳点和孤立点得以剔除。

Considering the shortcoming of least-square trend surface filter for outlier detection of multibeam echosounding,the quasiaccurate detection of outliers has been proposed to apply to detect the outliers of multibeam echosounding so that the method is combined with the fitting method for least-squares trend surface filter. By constructing trend surface of the different rank and analyzing the real data,the results show that the seabed terrain is relatively smoothing and the small terrain is reserved after deleting the outliers. In addition,the jump-points and isolated points are deleted.