A Three-dimensional Building Modeling Method Study Base On The LiDAR And DLG

In recent years,Airborne light detection and ranging(LiDAR)is a new detection technology over the ground,which has high spatial resolution and high time resolution.Using airborne light detection and ranging(LiDAR)point cloud data for building reconstruction,is widely used in urban construction,planning and dynamic monitoring,etc.So far,it is still difficult to applied in rapid modeling in large areas.At present,the manual measurement,remote sensing image and LiDAR point cloud are the main data sources of building reconstruction,and the use of different data sources and its accuracy will largely determine the precision of the building model and modeling time.In this study,on the basis of previous studies,the hotspot problem such as the filtering of LiDAR point cloud data and building information extraction were studied.The writer put forward a set of fast,efficient modeling method by using point cloud data.The main contents and achievements of this paper are as follows:(1)In the process of eliminating noise of point cloud data,we build height-frequency histogram,and normalize and smooth the histogram.After finding the peak of histogram,an iterative method is used to confirm the threshold of high and low noise point.Experiments show that,this method can effective eliminate the high and low point cloud noise,which has an effect on modeling filtering precision.During the filtering process of point cloud data,on the basis of typical filtering algorithm and comprehensive statistics related theory,an skewness balance algorithm was established,which is less manual intervention and nonthreshold selection.To the question that original skewness balance algorithm is not suitable to the area of topographic relief,we improve the algorithm and come up with the skewness balance algorithm based on altitude difference.Firstly,to divide grid of point cloud data,then to extract seed point and build initial ground TIN.Secondly,to calculate the altitude difference of point cloud and TIN,and to iterative compute the skewness of altitude difference.When coefficient of skew is not equal to zero,eliminate the highest point till skewness is equal to 0.The experimental results show that this method is appropriate for both flat terrain and rolling terrain.Besides,the filtering process doesn't need threshold.This process can be efficiently used in project practice of large area.(2)In building information extraction,we integrate digital line graph(DLG)and point cloud data by using the GIS overlay analysis method for building information extraction.To handle the weakness of DLG's timeliness,remote sensing image is used in densifying and checking the buildings of DLG.After intersect analysis,we build data buffer inside the building footprint to eliminate the points which is vegetation points or building side point.Experiment show that DLG can accurately extract building points from the point cloud data after filtering.Compared with other building information methods,DLG provides higher accuracy of building footprint.and higher extraction efficiency of building points.(3)During the process of building reconstruction,RANSAC algorithm is used to extract and fitting the roof patch.Then we extract the building contour by Alpha Shape operator.At last,we extract roof feature points by analysis of topological relations between patches.Combining with DLG building corner plane coordinates,we backcalculation space coordinate of top building contour points.The DTM which is interpolated by ground points is used to calculate the building bottom contour points,thus the building construction is completed.(4)We has carried on the building reconstruction of experimental area.We comprehensively evaluate the efficiency of this research's method by calculating reconstruction completion rate after vacuating the point cloud data.The results show that after several vacuations,the method can still maintain fine completion rate of model reconstruction.This method can be used in fast calculation of building plot ratio of testing zone.The results show that the research can be used widely in city planning,intensive land-use and other aspects.