Research On The Application Of Research On Airborne LIDAR Point Cloud Data Classification Method

As an emerging technology,airborne LIDAR technology has been applied in many industries because of its ability to acquire three-dimensional coordinate information on the ground quickly and efficiently.At present,research on LIDAR technology at home and abroad is relatively mature,but its data processing is still a hot research area.As the most important part of the data post-processing process,the results of data filtering will directly affect the quality of DEM generated afterwards.Therefore,it is an urgent problem to obtain LIDAR point cloud data quickly and accurately.So this paper was aimed to discuss two aspects:filter classification and point cloud thinning.Here was the main content:(1)This paper explained the composition of the airborne laser radar system and the working principle of laser ranging,clarified the system error of the airborne laser radar,and introduced the characteristics and application fields of the airborne laser radar system.The filter classification principle of airborne LIDAR data was discussed in this paper in detail.Meanwhile,the advantages and disadvantages of several commonly used filter algorithms at home and abroad are analyzed and summarized.(2)The traditional slope filtering algorithm is difficult to define when the slope of the terrain itself is not significantly different from that of the ground feature.An improved algorithm was proposed in this paper.An irregular triangulation mesh was used to simulate its irregular triangle surface.By judging the relationship between the undetermined point and its corresponding triangular facet,a threshold height difference was set,which effectively reduced the errors in Type II,and solved the self-adaptation problem of the slope threshold to a certain extent.(3)Using two existing LIDAR point cloud data algorithms of rarefying three kinds of different terrain point cloud data were experimentally studied.Finally,the relationship between the different rarefying rates and DEM accuracy in the two algorithms was obtained,After comparative and analysis.it was found that the accuracy of the DEM generated by point cloud data processing is still relatively good even if the 60%extraction rate is used.The research showed that the improved algorithm in this paper can effectively process the airborne LIDAR data,and deal with the similarity between the terrain slope and the slope of ground feature.And finally improve the accuracy of the filtering results.On the basis of meeting the requirements of the DEM practical engineering application accuracy,it is feasible to rarefy the filtered point cloud data.