论文标题
位置准确性评估历史谷歌地球图像
Positional Accuracy Assessment of Historical Google Earth Imagery
论文作者
论文摘要
Google Earth是当今最受欢迎的虚拟地球仪。鉴于其受欢迎程度和有用性,大多数用户都不会密切关注图像的位置准确性,并且有关该主题的信息有限。这项研究评估了2000年至2018年之间四个时期的历史GE图像的水平准确性,以及其在西非尼日利亚的拉各斯州内的高程数据的垂直准确性。通过与非常高的分辨率(VHR)数字矫正器进行比较,评估图像的水平精度,同时通过与558个地面控制点的网络进行比较评估垂直精度。还将GE高程与来自两个随时可用的30m数字高程模型(DEMS),Shuttle Radar地形任务(SRTM)v3.0和高级土地观察卫星世界3D(AW3D)DEM v2.1的高程数据进行了比较。最近的GE图像(2018年)是最准确的,而2000年是最不准确的。这显示了构成Google Earth数据来源的卫星图像数据源的准确性和可靠性的持续增强。就垂直精度而言,GE高程数据的最高RMSE为6213m,其次是AW3D,RMSE为4.388m,SRTM的RMSE为3682m。尽管SRTM和AW3D的垂直准确性优越,但Google Earth仍然在其易用性和上下文意识方面具有明显的优势。
Google Earth is the most popular virtual globe in use today. Given its popularity and usefulness, most users do not pay close attention to the positional accuracy of the imagery, and there is limited information on the subject. This study evaluates the horizontal accuracy of historical GE imagery at four epochs between year 2000 and 2018, and the vertical accuracy of its elevation data within Lagos State in Nigeria, West Africa. The horizontal accuracies of the images were evaluated by comparison with a very high resolution (VHR) digital orthophoto while the vertical accuracy was assessed by comparison with a network of 558 ground control points. The GE elevations were also compared to elevation data from two readily available 30m digital elevation models (DEMs), the Shuttle Radar Topography Mission (SRTM) v3.0 and the Advanced Land Observing Satellite World 3D (AW3D) DEM v2.1. The most recent GE imagery (year 2018) was the most accurate while year 2000 was the least accurate. This shows a continuous enhancement in the accuracy and reliability of satellite imagery data sources which form the source of Google Earth data. In terms of the vertical accuracy, GE elevation data had the highest RMSE of 6.213m followed by AW3D with an RMSE of 4.388m and SRTM with an RMSE of 3.682m. Although the vertical accuracy of SRTM and AW3D are superior, Google Earth still presents clear advantages in terms of its ease of use and contextual awareness.