基于RapidEye遥感影像的黄冕镇桉树时空动态分析

Spatiotemporal Dynamic Analysis of Eucalyptus in Huangmian Town Based on RapidEye Remote Sensing Images

文章基于RapidEye遥感影像,通过ENVI、ArcMap软件,使用随机森林、最大似然法、支持向量机、CART的方法获取2010年、2012年、2014年、2016年、2018年桉树人工林时空动态格局变化,并利用Google Earth Pro历史影像对实验结果进行了精度验证,结果表明:利用随机森林法提取黄冕镇桉树种植面积效果最佳,分类精度最高达到94.38%。2010年至2018年黄冕镇桉树种植面积逐年增加。研究结果可为日后桉树人工林的监测及种植提供科学依据。

Based on the RapidEye remote sensing image, the spatiotemporal dynamic pattern changes of eucalyptus plantations in 2010, 2012, 2014, 2016 and 2018 were obtained by using the methods of random forest, maximum likelihood, support vector machine and CART through ENVI and ArcMap software. The accuracy of the experimental results was verified by using the historical image of Google Earth Pro. The results showed that the effect of extractup toing eucalyptus planting area in Huangmian town by using the random forest method was the best. The highest classification accuracy reached 94.38%. From 2010 to 2018, the planting area of eucalyptus in Huangmian town increased year by year. The results can provide scientific basis for monitoring and planting eucalyptus plantation in the future.