基于影像分割技术的LIDAR数据建筑物边缘提取

Building Edge Extraction from LIDAR Data Based on Images Segmentation

提出一种基于影像分割技术的LIDAR数据建筑物边缘提取方法,将LIDAR点云数据转化为DSM影像,采用改进的动态阈值分割算法对DSM影像进行分割,并将得到的图像进行边缘提取。试验表明,该方法不需要其他辅助数据,可以从LIDAR点云数据中提取建筑物边缘,为进一步的建筑物三维重建奠定基础。

3D geographical information can be captured quickly from LIDAR.Building extraction based on LIDAR data processing is currently a hot topic and difficulty.In this paper,a method for building edge extraction from LIDAR data using image segmentation technology is proposed.The LIDAR data is first translated into DSM image.Then the improved dynamic threshold algorithm is used to segment DSM image.Lastly,the building edge is extracted on the segmented image.The experimental results show that the method can extract edge of buildings without accessorial data.It provides the preparation for the further 3D reconstruction of buildings,which has a certain value.