基于k-均值聚类算法的无人机雪域识别测绘

Unmanned Aerial Vehicle Snow Recognition Surveying Based on k-Mean Clustering Algorithms

本文介绍了基于k-均值聚类算法无人机航测技术在雪域自动测绘中的应用,通过k-均值聚类算法完成正射影像中的雪域边界识别,区分无雪面域和雪覆盖面域。本文以JD项目中Riz、Piz-E、Piz-W三个实验区块为例,运用无人机在连续积雪和斑块积雪两种情况下进行了现场测绘实验。通过无人机正射影像与红-绿-蓝(RGB)影像、近红外-红-绿(NIRRG)影像、近红外-绿-蓝(NIRGB)影像分别对比分析,证实了在红、绿、蓝波段下均可以完成无人机航测影像雪域识别任务,也进一步说明了在任意波段下无人机均能完成雪域自动识别任务。

This paper introduces the application of UAV aerial survey technology in automatic snow mapping, aiming at identifying snow boundary and distinguishing snow-free terrain and snow-covered terrain by K-means clustering algorithm. Taking Riz, Piz-E and Piz-W test blocks in JD project as examples, field mapping experiments are carried out using UAV under continuous snow and patch snow conditions. It is proved that UAV can accomplish the task of automatic snow recognition in any band.