高铁跨海路由单物标水深图提取算法研究

Research on Single Echo Sounding Coordinate Extraction Algorithm of High-speed Railway Over Sea

为了解决从CAD电子水深图中精确提取单物标水深数据实测点位坐标困难等问题。以海图软件生成的CAD水深图为样本,对各种典型数字字符水深数据进行归纳统计,分析字符的“宽度、高度、旋转角度、插入点坐标”与“实测坐标”之间的数学关系,推导根据水深字符插入点坐标反算水深实测坐标的公式。通过计算机编程,使用文件数据流I/O读写、多线程并发、数组分块等技术,直接访问CAD的DXF交换格式文件,实现海量数据的快速读取、识别、分类,以及坐标推算和XYZ文件报表生成。之后,对各种典型和极端情况下的模拟水深样例进行检验,该算法坐标偏差为1~3 cm。在福厦高铁泉州湾跨海大桥工程中,需要进行“回淤分析”“冲刷分析”等精细化分析计算,要求提取水深数据的位置误差<50 cm,常规人工提取方式无法满足精度要求,而使用该算法得到的水深坐标与原始实测坐标差值<5 cm。

In order to solve the problem that it is difficult to accurately extract the measured point coordinates of single object depth data from CAD electronic bathymetric chart,based on the statistical analysis of various typical numerical character water depth data examples,the mathematical relationship between width,height,rotation Angle,insertion point coordinates and measured coordinates of characters were found,and the formula of calculating measured coordinates of water depth according to the insertion point coordinates of water depth characters in CAD was derived.Through computer programming,using file data stream I/O reading and writing,multi-threading concurrency,array segmentation and other technologies,CAD DXF exchange format files were accessed,the rapid reading of massive data,identification,classification,coordinate calculation and the generation of XYZ file reports were achieved.The simulated water depth samples in various typical and extreme cases were tested by the proposed algorithm,and the statistical coordinate deviation was 1~3 cm.In the Quanzhou Bay cross-sea bridge project of the Fuzhou-Xiamen High-speed Railway,it was necessary to carry out fine analysis and calculation such as“silt back analysis”and“erosion analysis”,and it was required to extract the position error of water depth data<50 cm,the conventional manual extraction method could not meet the requirements.the algorithm in this paper was applied in practice,and the statistical difference between the extracted water depth coordinates and the original measured coordinates was less than 5 cm.