Intelligent Expression And Urban Change Detection Method Based On Multi-temporal LiDAR Data

With the rapid development of society and economy,the city due to human activities is constantly changing all the time.The continuous changes of cities results from a large number of new districts,old city reforming,the demolition of illegal buildings,roads change and expansion,newly built highways and railways,land reclamation of rivers and lakes and so on.In order to better understand the dynamics of urban development and provide support for urban planning,after fully concluding and analyzing the previous efforts,this thesis carried out relevant researches on key preprocessing technologies,intelligent expression based on spatial attributes and semantic features,The main research contents include the following aspects:(1)The key preprocessing technology airborne LiDAR data.This thesis carries out the research on point cloud strip adjustment model and filtering method in order to achieve high reliability of point cloud data.The error sources and difference mathematical adjustment model of airborne LiDAR system is systematically analyzed.Based on the mass data and the reliability of the model,an algorithm based on the point cloud adjustment model based on automatic connection line is proposed.Combining feature attributes,a multi-level filtering classification method based on feature attributes is proposed.(2)Intelligent expression of point cloud data based on spatial attributes and semantic features.This thesis studied the characteristics of different organization methods of point cloud data.Then,this thesis proposed a structured and semantic organization method of point cloud data fully exploring the spatial attribute,feature attribute,color attribute and topology attribute.In order to effectively and effectively carry out large scale point cloud data scheduling,this thesis carried out the research on the modularized TIN data structure designing,spatial index method,and triangulation network simplification calculation.(3)Urban change detection method based on multi-temporal LiDAR point cloud.Considering the characteristics of spatial property and features of modern cities,this thesis carried out the change detection of urban terrain and construction.In order to detect the changes of terrain,this thesis proposed a method for three dimensional terrain change detection based on improved ICP and slope entropy downsampling algorithm.In order to detect changes of construction,this thesis proposed a method for three dimensional change detection based on morphological operator.