基于遥感技术的镇沅县茶树资源研究  

Research on Tea Resources in Zhenyuan County Based on Remote Sensing Technology

作　　者：段宗亮 周晓 DUAN Zongliang;ZHOU Xiao(Yunnan Academy of Forestry and Grassland,Kunming Yunnan 650201,P.R.China)

机构地区：[1]云南省林业和草原科学院,云南昆明650201

出　　处：《西部林业科学》2024年第5期140-145,共6页Journal of West China Forestry Science

基　　金：国家自然科学基金(32260105)。

摘　　要：为提升茶树种质资源调查的效率和准确性,本研究基于全球尺度卫星数据量大、计算分析能力强的谷歌地球引擎(GEE)平台,以镇沅县为研究区域,利用哨兵2号(Sentinel-2)影像的光谱、植被指数、纹理特征,综合运用随机森林算法(RF)、支持向量机(SVM)和网络模型(BP)神经网络算法实现了茶树资源的提取。该方法和思路首次应用于茶树资源调查,结果显示:(1)基于RF算法的茶树资源提取效果和精度最佳,总体精度为85.6%,说明RF算法在监测和提取茶树资源方面具有较大的应用价值和前景。(2)镇沅县茶斑块3 430块,茶树资源面积约1.9×10^(4) hm^(2),其中天然茶树资源面积1.15×10^(4)hm^(2),人工茶树资源面积0.75×10^(4)hm^(2)。天然茶树面积占60.5%,人工茶树面积占39.5%。在空间分布上,天然茶树分布于较陡山地,人工茶树都位于较为平缓的村庄周围,茶资源区划准确,效率高。To improve the efficiency and accuracy of tea tree resource investigation,this study is based on the GEE platform,which has a large amount of global satellite data and strong computing and analysis capabilities.Taking Zhenyuan County as the research area,the spectral,vegetation index,and texture features of Sentinel-2 images were used to extract tea tree resources,and RF,SVM,and BP neural network algorithms were comprehensively applied.This method and approach were first applied to the investigation of tea tree resources,and the results showed that:the tea tree resource extraction effect and accuracy based on RF algorithm were the best,with an overall accuracy of 85.6%,indicating that RF algorithm has great application value and prospects in monitoring and extracting tea tree resources.The research results indicate that the tea resource area in Zhenyuan County is about 1.9×10^(4) hm^(2),with 3430 tea patches.Among them,there are 1.15×10^(4) hm^(2) of natural tea resources;Artificial tea covers an area of 0.75×10^(4) hm^(2).Natural tea tree resources have a large distribution area,while artificial tea trees are widely distributed,with about 80%of villages having them.So accurate tea resource zoning and high efficiency.

关 键 词：GEE平台 遥感技术 分类方法 茶资源 

分 类 号：S771.8[农业科学—森林工程]