首页|老龄台地茶园遥感信息提取方法研究——以西双版纳州为例

老龄台地茶园遥感信息提取方法研究——以西双版纳州为例

Study on Remote Sensing Information Extraction Method of Aged Tea Garden:Taking Xishuangbanna as an Example

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
为利用遥感技术探测老龄台地茶园信息,根据茶园物候特征,基于研究区Sentinel-2、GF-2、LANDSET5-TM遥感影像数据与30 m分辨率的DEM等辅助数据,结合植被指数、光谱特征、纹理特征、地形特征,采用随机森林分析方法提取30年老龄台地茶园空间分布与茶园面积信息,并结合官方统计数据对研究结果进行精度验证.结果表明:①1990年茶园提取结果的Kappa系数达到0.83,总体精度OA达到88.5%,茶园的生产精度PA精度为83.73%,UA为89.68%;2020年茶园提取结果的Kappa系数达到0.80,总体精度OA达87.13%,茶园的生产精度PA为84.18%,用户精度UA为94.10%.②使用随机森林分类方法并结合光谱和纹理等特征提取的茶园精度较高,符合生产需求.③1990年全州台地茶园面积为15 260 hm2,2020年台地茶园面积为96 986.67 hm2.遥感技术可用于探测老龄台地茶园的空间分布与种植面积,为低产茶园改造提供数据支持与调控参考.
In order to identify information about old tableland tea gardens using remote sensing technology,we employed a random forest analysis based on Sentinel-2,GF-2,LANDSAT 5-TM remote sensing image data,along with 30-meter resolution Digital Elevation Model(DEM)and other auxiliary data in the study area.This analysis enabled us to extract the spatial distribution and area of tea gardens that are over 30 years old.The process involved utilizing vegetation indices,spectral features,texture characteristics,and topographic attributes to enhance the accuracy of the results.The results showed that:① the Kappa coefficient of the extracted results of 1990 tea plantations reached 0.83,the overall precision OA reached 88.5%,the production precision PA of tea plantations was 83.73%,UA was 89.68%;the Kappa coefficient of the extracted results of 2020 tea plantations reached 0.80.The overall precision OA reached 87.13%,the production precision PA of tea gardens was 84.18%,and the user precision UA was 94.10%.② The precision of the information of tea gardens,which was extracted using random forest classification method and incorporating spectral and textural features,demonstrated elevated precision,perfectly aligning with production requirements.③ The area of tea gardens on the tablelands in the state was 15260 hm2 in 1990,and the area of tea gardens on the tablelands in 2020 was 96986.67 hm2.The remote sensing technology can be used to detect the spatial distribution and planting area of the old tableland tea plantations,as well as provide data support and regulation reference for the transformation of low-yield tea plantations.

satellite remote sensingrandom foresttea plantation distributionolder tea treesinformation extractionvegetation indexspectral characteristicstexture features

龙玉兰、张军、唐莹、刘逸娴

展开 >

云南大学国际河流与生态安全研究院,昆明 650504

云南大学地球科学学院,昆明 650504

卫星遥感 随机森林 茶园分布 老龄茶树 信息提取 植被指数 光谱特性 纹理特点

国家国防科技工业局高分专项云南省政府综合治理深度应用与规模化产业化示范项目子课题云南大学研究生创新人才培养项目

89-Y50G31-9001-22/23cz22622202

2023

10.12396/jsgg.2022409

节水灌溉
中国国家灌溉排水委员会,中国灌溉排水发展中心,武汉大学,国家节水灌溉北京工程技术研究中心

节水灌溉

CSTPCD北大核心
影响因子：0.674
ISSN：1007-4929
年,卷(期)：2023.(12)
  • 1
  • 20