Title

UAV航拍技術應用於河道變遷土砂監測和山區地形製圖之可行性分析

Translated Titles

Feasibility Analyses of River Sediment Estimation and Mountain Mapping by UAV Techniques

Authors

張崴(Wei Chang);蕭宇伸(Yu-Shen Hsiao);張榮傑(Jung-Chieh Chang)

Key Words

無人飛行載具 ; 數值地形模型 ; 精度分析 ; UAV ; DSM ; accuracy analysis

PublicationName

中華水土保持學報

Volume or Term/Year and Month of Publication

48卷1期(2017 / 03 / 01)

Page #

1 - 13

Content Language

繁體中文

Chinese Abstract

本研究利用無人飛行載具搭載相機,配合高精度之航測控制點,於南投縣貓羅溪和惠蓀林場進行空拍,以影像處理軟體Pix4Dmapper 建立該區域之正射鑲嵌影像與數值表面模型。本研究採用旋翼機獲取貓羅溪空拍結果後,計算一年期間 (2015 年4 月至2016 年4 月) 之泥沙變化量,並利用地面檢核點分析平面與垂直方向之誤差。惠蓀林場則使用定翼機進行空拍,配合不同分布情形之地面控制點進行平面與垂直方向之誤差分析。本研究主要相關結果如下:(1) 在使用四旋翼UAV 搭載GoPro 錄影式相機建立之河道數值表面模型在X 方向精度為9.2 公分、Y 方向精度為5.3 公分、Z 方向精度為4.8 公分;(2) 在使用六旋翼UAV 搭載Sony NEX7相機建立之河道DSM 在X 方向精度為3 公分、Y 方向精度為4.1 公分、Z 方向精度為1.6 公分;(3) 利用前後期數值表面模型計算河道泥沙變化量為226,971.49 立方公尺,每1 立方公尺估計量之內部精度為0.202 立方公尺;(4) 在惠蓀林場方面,將控制點均勻分布之地形精度明顯優於只將控制點分布在地勢較低之區域。

English Abstract

We use unmanned aerial vehicles (UAV) with high-accuracy control points to obtain orthophoto mosaics and digital surface models (DSM) by Pix4Dmapper in parts of the Mao-Luo stream and Hui-Sun forestry areas, Nantou County. Two different rotor UAVs were used to carry out the missions in the Mao-Luo stream area. The results were used to estimate the river channel sediment variation from April 2015 to April 2016, and were also evaluated with ground check points with both horizontal and vertical components. A fixed-wing UAV was used for the mission in the Hui-Sun forestry area. The accuracies of the results using different distributions of control points were also computed and analyzed with both horizontal and vertical components. The main results of this study include: (1) The DSM derived from the 4 rotor UAV with a GoPro yields the accuracies of 9.2, 5.3, 4.8 cm at directions of X, Y, and Z; (2) The DSM derived from the 6 rotor UAV with a Sony NEX7 camera yields the accuracies of 3.0, 4.1, 1.6 cm at directions of X, Y, and Z; (3) The river channel sediment variation from April 2015 to April 2016 is 225,685.61 m^3, and the internal accuracy for 1 m^3 earthwork reaches 0.202 m^3; (4) In Hui- Sun forestry area, the accuracy of UAV-derived DSM using control points distributed averagely is obviously better than that using control points distributed over low elevations only.

Topic Category 生物農學 > 農業
生物農學 > 森林
生物農學 > 畜牧
生物農學 > 漁業
生物農學 > 生物環境與多樣性
工程學 > 土木與建築工程
工程學 > 市政與環境工程