Several Key Technologies Of Airborne LiDAR Based Provincial Foundation Mapping

Airborne LiDAR (Light Detection and Ranging) is new type observation techology from space to earth. It is capability of obtaining rapidly data high precise, wide area, high density sampling, three dimensional coordinates, even including certain penetrability, many regain pulse and it’s intensity information, spectrum image and so on. It is a progress of traditional Photogrammetry and Remote sensing. It solved the difficult problem of setting high precise digital elevation model on mountain area, tidal flat, the desert, forest, island and so on. Undoubtedly airborne LiDAR is a milestone of surveying and mapping technology development.Because of the diversity of units of airborne LiDAR system and the complex movement of the installation platform at high altitude, the airborne LiDAR data are affected by many error sources, so that raw data include random errors, gross errors and systematic errors. In addition, airborne LiDAR system adopt semi-random mode to sample, so it hasn’t classify label to sampling point. This is why it must be error processing and classification corresponding before using. Of course it also is one of most difficult problem of airborne LiDAR on the engineering practice at present. Based on carefully analysis airborne LiDAR scanning data, the dissertation combine the project of Jiangsu province foundation surveying and mapping elevation renew to investigate a few key technologies of the algorithm to remove the gross error point of raw data, to classify point clouds into the bare earth point and feature point, to compentate system error of the point clouds. Making the processed data is capable of meetting elevation renew request of province level. In addition, in order to perfect techology of province level surveying, boatborne LiDAR system of the new technology on the sea is also simplily investigated in the dissertation.The dissertation main studies are described as follows:(1) The dissertation introduces airborne LiDAR system composition and working principle briefly, and analyzes error sources of airborne LiDAR data. It also discusses point clouds data characteristic and classical gross point eliminating method, filtering algorithm of the bare earth point and feature point, systematic error testing method and eliminating strategy.(2)Eliminating low gross point Based on fitting plane.The method and flow to eliminate low gross point of airborne LiDAR point clouds data are studied in the dissertation. Aimming at elevation saltation condition in local region, a sort of method and flow to eliminate low gross point based on least square fitting plane are put forward based on sum up and analysis to eliminate method of the gross point in existence. Firstly, the object point is taken as center to build a unit cell, then in order to eliminate higher elevation feature points, the laser points in the unit cell are preprocessed, and then the rest laser points are used to adopt the least square principle to fit a plane. Finally, according to the distance between the object point and the plane judgement whether it is a low gross point. Using fitting plane method eliminate low gross points of four samples data of airborne LiDAR supplying by ISPRS group3, and qualitative analysis the experiment results.(3) Filtering algorithm based on multi-region growth.Combining point clouds data characteristic and realistically relief variety, the dissertation study the method which classify the point clouds into bare earth point and feature point. Based on analysis conventional filtering algorithms at present, the dissertation put forward one kind of fitting algorithm and flow which based on multi-region growth to classify bare earth point and object point. Firstly, segmentation the whole area into regular small rectangle region (window), then the lowest point is regarded as seed point of bare earth in window, it compares with other laser point which its’neighbors according to threshold of slope to distinguish whether the object point is a bare earth point. Continually compare with other object points of its’neighbors region as such to end. Tag as bare earth point do growth in window continually, until have not new bare earth point growth. Using such means do grow the bare earth point to other window, until the growth of bare earth over the whole area. Finally, the algorithm anew segmentate the whole area into different window. Tag as bare earth point do growth in window continually, such can classify the laser point clouds into the bare earth point and feature point. Multi-region growth filtering algorithm is used to classify the point clouds into the bare earth point and feature point which six samples data of airborne LiDAR supplying by ISPRS group3and two samples data sumplying by Jiangsu province surveying and mapping bureau, then quantitative analysis and qualitative analysis to the experiment results are conducted.(4) Systematic error compensating method based on least squares collocation.The dissertation studies the compensating method of airborne LiDARdata systematic error. Based on sum up and analysis conventional methods of eliminating systematic error, according to "data driven" theory, the dissertation put forwards one kind systematic error compensating method which considering random error and tendency parameter of the least squares collocation. Firstly, using network RTK survey a number of points which covering the whole test area, and obtaining elevation value of points which comparatively position. Then statistics test was performed to determine whether there exists the systematic error. The systematic error of existence will be considered one tendency parameter. Using the least squares collocation set a mathematics model to estimate the tendency parameter and adjusted value of test points. According to the tendency parameter compensate systematic error of all points in test area. Finally, Based on airborne LiDAR design province surveying elevation renew technology scheme.(5) Island surveying based on boatborne LiDAR.In the dissertation boatborne LiDAR is studied to survey island. On the basis of the preliminary study, the ship features, LiDAR measurement system unit composition and main function, the algorithm flow are introduced. At the same time, the scanning point clouds data before and after correction is conducted qualitative analysis. Lastly, after thin points, eliminate gross error points, classify the bare earth point and feature point and so on, boatborne LiDAR data which belong to two island of Yellow Sea is used to make contour maps,...