基于数字图像处理技术的LiDAR数据建筑物提取  

Building extraction from LiDAR data based on digital image processing

作　　者：唐咸远[1] 周国恩[1] 梁鑫[1] 

机构地区：[1]广西工学院土建系,广西柳州545006

出　　处：《四川建筑科学研究》2010年第3期335-337,共3页Sichuan Building Science

基　　金：广西自然科学基金资助项目(桂科自0728045)

摘　　要：LiDAR技术已经被广泛地应用于城市三维模型的重建。本文在现有方法的基础上,探讨了一种基于数字图像分析技术提取建筑物边界的方法,该方法首先将LiDAR点云数据量化生成灰度影像,然后利用图像分割、边缘检测等图像分析技术将建筑物信息与地面特征分离开,最终提取建筑物的轮廓线。实验结果表明,这种方法能够比较准确地反映建筑物的边界,效果理想。LiDAR data had been widely used in 3D building construction.The paper discussed a method for building extraction from LiDAR data based on digital image processing.Firstly,the LiDAR point cloud data was quantified by 256 degree and then translated into gray image;secondly,the image segmentation and edge detection algorithm was applied for edge extraction of buildings.The test result showed that the method could extract building exactly,and had great reference in practice.

关 键 词：激光扫描测距 数字图像处理 图像分割 边缘检测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]