Building Boundary Extraction Using LiDAR Pointcloud And CCD Images

With the development of information technology, digital city construction is in full swing,and the buildings are the main element of digital city, so building boundary extraction is supposed to the key technology in building modeling. The development of LiDAR brings data acquiring a lot of convenience, and the information acquired is very abundant? According to the superiority of LiDAR data which has 3D information and images data which has rich texture information, a method that extracting building boundary using LiDAR point cloud data and CCD images has been put forward in this article, which can extracting building boundary self-motion and Higher precision. The research contents are as follows:Pretreatment of the raw data, culling, filtering and extraction the building points three process, including errors point to prepare in advance, after a short walk to the back of the building point pretreatment good work.Clustering method regional growth point cloud of groups of buildings divided between the buildings and the buildings can not be destroyed, there are errors and other reasons the process of excluding non-building data points, each building is divided into separate buildings of the existing ray method to improve the use of artificial the error segmentation buildings again the visual judgments segmentation.Create building images -after creating a buffer and a bounding rectangle of a building according the segmented building points, filter image using the buffer and then cut the result image by the bounding rectangle.Line segments extraction-most buildings have perpendicular structure with two rectilinear axes. So an automatic and robust method on building principal orientations estimation is presented based on rough principal orientations constraint, thus improving the accuracy and robustness of line segments extraction.Boundary selection -a selection strategy is proposed based on LiDAR data density analysis and K-means clustering.