本文關(guān)鍵詞： 機載LiDAR 點云 濾波 分類　出處：《中國礦業(yè)大學(xué)（北京）》2013年博士論文　論文類型：學(xué)位論文

【摘要】：機載LiDAR系統(tǒng)是一項新興的測繪技術(shù)。論文系統(tǒng)總結(jié)和分析了機載LiDAR系統(tǒng)的測距原理�；贚iDAR點云的高程直方圖及點云間的相互關(guān)系濾除了點云噪聲，提出了分類LiDAR地面點云的三種算法：改進參數(shù)的形態(tài)學(xué)法、二元二次趨勢面結(jié)合形態(tài)學(xué)法、窗口迭代的克里金法，經(jīng)過試驗數(shù)據(jù)驗證知，三種方法均取得了較好的分類結(jié)果。利用CUMT_PC法和CUMT_AT的方法分別分類出建筑物點云，并提取出了屋頂面域，試驗結(jié)果表明兩種方法的分類精度可以達到中等水平，，且屋頂面域的探測率和質(zhì)量因子都高于92%，能達到自動識別的要求。根據(jù)點云的高程和反射強度信息分類出城市中的道路點云，提出了一種城市主干路網(wǎng)的提取算法，提取的城市道路網(wǎng)與參考數(shù)據(jù)對比后發(fā)現(xiàn)，所提取道路網(wǎng)的完整率和正確率較高。
[Abstract]:Airborne LiDAR system is a new technology of surveying and mapping. The principle of airborne LiDAR system ranging is summarized and analyzed in this paper. The elevation histogram based on LiDAR point cloud and the relationship between point clouds are filtered. Except point cloud noise. Three algorithms for classifying LiDAR ground point clouds are proposed: morphological method of improved parameters, binary quadratic trend surface combined with morphology method, window iterative Kriging method, and verified by experimental data. The CUMT_PC method and the CUMT_AT method are used to classify the building point cloud, and the roof area is extracted. The experimental results show that the classification accuracy of the two methods can reach medium level, and the detection rate and mass factor of the roof area are both higher than 92%. According to the elevation and reflection intensity information of the point cloud, the road point cloud in the city is classified, and an algorithm for extracting the urban trunk road network is proposed. Comparing the extracted urban road network with the reference data, it is found that the integrity and accuracy of the extracted road network are higher.
【學(xué)位授予單位】：中國礦業(yè)大學(xué)（北京）
【學(xué)位級別】：博士
【學(xué)位授予年份】：2013
【分類號】：P225.2

【引證文獻】

相關(guān)碩士學(xué)位論文 前1條

1 林鑒;機載LIDAR點云數(shù)據(jù)濾波及建筑物點群分割研究[D];西南交通大學(xué);2014年

本文編號：1490931