Study On Extraction Of Glacier Snow Line Based On Airborne LiDAR In Qinghai-Tibet Plateau

The snow-capped mountains and glaciers on the Qinghai-Tibet Plateau are the source of many river systems,directly supporting the sustainable use of freshwater resources in China.Studying the boundary changes of glaciers and snow-capped mountains can effectively grasp the local climate change rules,which has a profound impact on the development of meteorology.How to extract the boundary of glacier snow line reasonably and effectively has important practical and practical significance.With airborne laser radar data collection is dense,strong timeliness,high precision,high speed,and the advantages of short operation period.This article selects some typical Qinghai-Tibet mountain area of high density,high precision airborne Lidar point cloud data of ice snow line contour line extraction,so as to achieve the accuracy of the dynamic change law study late ice snow line.The core research content of this paper can be summarized into the following three aspects:(1)According to the topographic characteristics of the study area,a step-by-step itemized denoising algorithm for point cloud is proposed.Firstly,the outer noise is removed by using the theory of limit error in measurement error,and then the inner noise is removed by meshing and quadric surface fitting.Through experimental analysis and comparison,it is concluded that this method can effectively eliminate the noise point cloud and has certain feasibility and reliability.(2)This article research the filtering algorithm suitable for point cloud data in mountainous areas of the Qinghai-Tibet Plateau.Point cloud data filtering based on encrypting of irregular triangle net(TIN),through to the ground point cloud build irregular triangle net(TIN),within the study area on each filter through a LiDAR point cloud,continuously,continuously iteration,thereby to determine choice,until all the airborne LiDAR point cloud all determine screening is completed,which will ground point and the ground point for effective separation;The statistical analysis and comparison of the filtering errors in the study area are carried out by qualitative and quantitative methods.The results show that the filtering algorithm in this paper has a better filtering effect,which is better than that of lidar360,and well preserves the topographic and geomorphic information of the study area.(3)In the strength of airborne LiDAR point cloud data information based on the analysis of the characteristics,combined with the snow-capped mountains of the point cloud echo intensity is much higher than other mountains area feature this characteristic,this paper puts forward the point cloud data clustering,the strength of the information of existing K-MEANS clustering algorithm was improved,the algorithm in the thought of traditional Euclidean distance to join matrix,improvement for markov distance,very good to avoid the interference of the correlation between variables,using the distance evaluation function(DEF)a standard to choose one of the most appropriate and most perfect way of clustering.Through experimental verification,analysis and comparison,it is shown that the improved k-means clustering algorithm can achieve better clustering effect.The contour lines of some snow-capped mountains can be extracted by using lidar360 point cloud data processing software,and the elevation information can be effectively displayed.