西湖龙井茶园卫星遥感识别研究  

Research on Satellite Remote Sensing Identification of Xihu Longjing Tea Plantation

作　　者：杨军 范辽生 YANG Jun;FAN Liaosheng(Hangzhou Meteorological Information Center,Hangzhou 310051,china)

机构地区：[1]杭州市气象信息中心,浙江杭州310051

出　　处：《中国茶叶》2023年第9期32-36,共5页China Tea

摘　　要：文章基于吉林一号高分辨率卫星以及中低分辨率卫星资料,利用实例分割法和改进的HRNetV2模型,对西湖龙井茶园进行初步自动提取,在此基础上结合人工解译,进一步获取精细空间分布和面积统计。通过新HRNetV2模型识别出的茶园边界与居住区以及山上林木边界清晰,纹理细腻丰富,茶园面积识别准确率接近92%。研究对西湖龙井品牌保护、数字化管理、精细化气象服务,以及茶树生长环境小气候研究具有重要意义。The article was based on the high-resolution satellite data of Jilin-1 and the medium to low resolution satellite data.Using the instance segmentation method and the improved new HRNetV2 model,a preliminary automatic extraction of the Xihu Longjing tea gardens were carried out.On this basis,combined with manual interpretation,accurate spatial distribution and area statistics were further obtained.The boundaries of tea gardens,residential areas,and trees on the mountains identified by the new HRNetV2 model were clear,with delicate and rich textures.The accuracy of tea garden area recognition was close to 92%.The research was of great significance for the protection of the Xihu Longjing brand,digital management,refined meteorological services,and microclimate research in the growth environment.

关 键 词：西湖龙井 茶园 吉林一号 新HRNetV2模型 

分 类 号：S571.1[农业科学—茶叶生产加工] S771.8[农业科学—作物学]