国际中文开源期刊平台

一种K-Means算法的坐标转换公共点自动选取方法

A K-Means Algorithm for Automatic Selection of Public Points for Coordinate Conversion

测绘观察 / 2021,3(1)：9-21 / 2021-02-24 600 842

查看全文 下载文档

• 作者： 刘国栋      秦浩      刘佳      刘浪     
• 单位：

  重庆交通大学土木工程学院，重庆

• 关键词： 坐标转换；重合点自动选取；K-Means 算法；Bursa 模型；Python
• Coincidence point selection; Coordinate conversion; K-Means; Bursa; Python
• 摘要： 背景：坐标转换中，在具有多个已知数据的重合点时，重合点的选取是关键问题之一。目标：一个基于 K-Means 算法在众多重合点中自动选出合适重合点的方法。方法：通过分析 K-Means 聚类算法的基本原理和最佳 k 值的确定，针对空间直角坐标数据提出了一个初始聚类中心的选取方法，并结合 Python 实现了从多个重合点中快速、自动地选取出重合点，同时使用 Bursa 模型解算出七参数和内符合精度达到要求。结果：利用 K-Means 聚类算法能够快速、自动地选出较为合适的重合点，内符合精度在最佳 k 值处最优，由此确定的重合点求解出的七参数能够用于坐标转换中。结论：研究表明基于 K-Means 算法自动选出重合点的方法不但可行而且能够满足坐标转换精度要求，对坐标转换工作具有非 常重要的意义。
• Background: In coordinate conversion, when there are multiple coincident points with known data, the selection of coincident points is one of the key issues. Objective: A method based on the K-Means algorithm to automatically select a suitable coincidence point among many coincidence points. Method: By analyzing the basic principles of the K-Means clustering algorithm and determining the optimal k value, an initial clustering center selection method is proposed for spatial rectangular coordinate data, and combined with Python to achieve rapid selection from multiple coincident points, the coincidence point is automatically selected, and the Bursa model is used to calculate the seven parameters and the internal compliance accuracy meets the requirements. Result: The K-Means clustering algorithm can quickly and automatically select a more suitable coincidence point, and the internal coincidence accuracy is the best at the best k value. The seven parameters solved by the determined coincidence point can be used in the coordinates Converting. Conclusion: The study shows that the method of automatically selecting coincident points based on the K-Means algorithm is not only feasible but also able to meet the accuracy requirements of coordinate conversion, which is of great significance to coordinate conversion work.
• DOI： https://doi.org/10.35534/go.0301002
• 引用： 刘国栋，秦浩，刘佳，等．一种 K-Means 算法的坐标转换公共点自动选取方法［J］．测绘观察， 2021，3（1）：9-21．