基于机载LiDAR的像控点自动提取方法研究

Research on Automatic Extraction Method of Image Control Points Based on Airborne LiDAR

铁路工程应用中,为解决机载LiDAR点云与影像平面位置精度不一致的问题,采用高精度点云分类方法,基于地面点与建筑物点分类数据,自动提取航飞范围内建筑物的角点坐标,为后续影像空三加密提供像控点成果。通过介绍地面点、植被点与建筑物点的分类方法、建筑物矢量边界自动生成与像控点坐标提取的方法,基于国外某铁路的GPS实测数据,对自动方法与传统手动方法提取的建筑物角点坐标进行了精度评价,经实验对比发现,自动方法提取像控点平面位置中误差为0.28 m,而传统手动方式平面位置中误差为0.39 m,相较于传统手动方法,自动方法精度提高了0.1 m左右,作业效率提高50%以上。在实际生产项目中,可显著提高内业数据处理效率。

In order to solve the problem of inconsistency between the LiDAR point cloud and the image plane position accuracy in the application of railway airborne LiDAR railway engineering,a high-precision point cloud classification method is adopted.Based on the classified ground point and building point data,the structure of the building within the flight range is automatically extracted.The corner point coordinate method provides image control point results for the subsequent image space three encryption.The thesis introduces in detail the classification methods of ground points,vegetation points and building points,the automatic generation of building vector boundaries and the extraction of image control point coordinates.The GPS measured data of a foreign railway was used to evaluate the accuracy of the automatic method and the traditional manual method to extract the corner point coordinates of the building.The experimental comparison and verification found that the automatic method extracts the plane position of the image control point with an error of 0.28 m.The manual method has an error of 0.39m in the plane position.Compared with the traditional manual method,the method of automatically extracting the control points o has increased the plane position accuracy by about 0.1 m,and the operation efficiency has increased by more than 50%.The actual production project has significantly improved the internal The efficiency of industrial data processing.