Extraction Of The Surface Features Based On Lidar And Remote Sensing Image Fusion

At the end of 80'of twenty century, Light Detection And Ranging(LIDAR) as a new Geo-spacial information technology turned out, it is an emerging and attractive technology that is capable of rapidly capturing three dimensional geo-spacial information in multi-scales, and also give us a new technology for obtaining short time resolution and high Spacial resolution data Data capturing is continuous,not manually taken point to point,which also make the automatic and intelilgent data processing feasible.LIDAR data turns out to be discrete "point cloud" in space.The clouds come from objects' surface which can return laser pulse.In order to rebuild Digital Elevation Model(DEM),we should wipe off non-ground points firstly,namely LIDAR digital filtering.The question have been studied for many years internationally.Some of the principle and algorithm of data filtering are discussed in the paper.A new filtering algorithm is brought forward based on research data,the result can be satisfied the need.In order to detect building region for reconstructing buildings, LIDAR data needed to be classfied,Namely LIDAR data classfy.Building model is an important application for LIDAR, also a difficulty for information extraction,in the paper,we put forward a method for detecting objects by fusing lidar data and images,export the building region for model reconstruct.A simple 3Dcity model turns out in the end.The main studies and contributions are described as follows:1.The principle and system structure of LIDAR are introduced.A certain applications such as mapping ,environmental monitoring,DEM generation ,as well as 3D city model are outlined.Great potential of LIDAR is proved through the advantages and disadvantages with respect to various aspects between lidar and photogrammetry.2.The common methods and algorithms to filter LIDAR data are reviewed in detail,the advantages and disadvantages are discussed.Then we use a new method for LIDAR data filtering,the result turns out well. 3.Study LIDAR data organization,use rectangle rule grid for data organization and analyse the common interpolation algorithms,then we use Kriging and inverse Distanee Weighted Intepolation algorithms for DTM and DSM.4.Common methods for classfying LIDAR are discussed in detail.we put forward a new method for detecting objects by fusing lidar data and remote sensing images,and evaluate the result finally.5.Export the building area,extract the edge and rebuild the plane buildings by useing the average height within the house in GIS software,finally,a simple 3D city model is present.The method proved to be a quick means for digital city model.