Analysis of Horizontal Accuracy for Large Scale Rural Mapping Using Rotary Wing UAV Image

5th Geoinformation Science Symposium 2017

地球科学;计算机科学

Hidayat, Husnul^1 ; Muljo Sukojo, Bangun^1

Department of Geomatics Engineering, ITS, Surabaya, Indonesia^1

关键词: accuracy;    Accuracy assessment;    Differential GPS;    Georeferencing;    Geospatial data sources;    Root mean square errors;    Structure from motion;    Very high resolution;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/98/1/012052/pdf
DOI  :  10.1088/1755-1315/98/1/012052

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

In order to fulfil the need of large scale map for rural mapping, Indonesian government is now looking for alternatives of geospatial data sources. With newly developed technology, nowadays rotary wing UAV can be used to acquire very high resolution aerial imagery quickly with low cost. This research assesses the horizontal accuracy of rural mapping in 1:2000 scale using orthophoto derived from rotary wing UAV image data. The test site of this research is Kebonwaris village, Pandaan, East Java which has an area of approximately 167 hectares. Image data was taken with approximately 80% overlap each other and processed using Structure from Motion approach. Twelve GCPs coordinates were measured using differential GPS observations for georeferencing purpose. For accuracy assessment, 22 test points were established and their coordinates were measured using static differential GPS observations. The results show that the mean absolute horizontal errors are 0.071 m and 0.142 m for easting and northing respectively and the Root Mean Square Errors are 0.088 m and 0.169 m for easting and northing respectively. These RMSE values represents horizontal RMSE 0.190 m. According to Peraturan Kepala BIG No. 15 Tahun 2014, this RMSE value represents the 0.289 m horizontal accuracy based on CE90 criterion. Therefore, with this level of accuracy the UAV image data can be used to make a class 1 base map in 1:2500 scale or class 2 base map in 1:1000 scale.