高分1号数据用于云南文山三七种植信息提取  

GF-1 remote sensing data for Panax notoginseng planting information extraction in Wenshan, Yunnan Province

作　　者：朱赞 袁希平 甘淑[1,3] 张良洁[1] 赵庆会 ZHU Zan;YUAN Xiping;GAN Shu;ZHANG Liangjie;ZHAO Qinghui(Faculty of Land Resources Engineering,Kunming University of Science and Technology,Kunming 650093,Yunnan,China;West Yunan University of Applied Sciences,Dali 671006,Yunan,China;Surveying and Mapping Geo-Informatics Technology Research Center on Plateau Mountains of Yunnan Higher Education,Kunming University of Science and Technology,Kunming 650093,Yunnan,China)

机构地区：[1]昆明理工大学国土资源工程学院,云南昆明650093 [2]滇西应用技术大学,云南大理671006 [3]昆明理工大学云南省高校高原山区空间信息测绘技术应用工程研究中心,云南昆明650093

出　　处：《浙江农林大学学报》2020年第1期129-135,共7页Journal of Zhejiang A&F University

基　　金：国家自然科学基金资助项目(41861054,41561083)

摘　　要：[目的] 云南文山三七Panax notoginseng作为经济价值较高的中草药,其产量多少与价格波动之间存在着较紧密的联系,实时准确地掌握三七种植面积,对于当地政府相关管理部门科学指导三七种植规模和确定价格等宏观调控具有重要意义。 [方法] 以文山州4个三七主产县为研究区域,基于国产16 m分辨率的高分1号(GF-1)开源影像,根据三七的生长环境特点对专家决策树模型进行改进,并通过决策树分类法提取三七的荫棚图斑。在对三七荫棚识别可靠度和面积提取精度评价中,采用了谷歌影像代替实地调绘和以目视解译结果作为图斑面积基准的方法。 [结果] 决策树分类成果和目视解译提取的图斑判定正确率分别为87%和99%;专家决策树分类提取的成果整体面积精度为80%。 [结论] 与传统基于高分辨率商业影像采用目视解译的人工勾绘图斑方法相比,基于GF-1影像的三七面积提取所采用的方法能够在保证一定精度的条件下,以较快速度开展文山三七种植资源的调查,为当地特色农产品种植的科学监管与价值估算提供基础数据与有效技术支持。Objective Panax notoginseng, a kind of Chinese herbal medicine with high economic value and a close relationship to price fluctuation, could be very meaningful for a government's agricultural management departments to scientifically guide its planting scale and determine the price of macro controls. The aim is to obtain P. notoginseng planting information in real time and accurately. Method Four counties which mainly produce P. notoginseng in Wenshan Prefecture were taken as the research area. According to the characteristics of P. notoginseng growth, environment construction expert decision tree, the decision tree classification method to extract P. notoginseng shade patches and a method of refining processing on classification results through GIS spatial analysis, was used. Based on these, the area of P. notoginseng from the Gaofen-1 (GF-1) remote sensing images with 16 m resolution was extracted, and the extracted results were refined. Then, Google image was used, instead of field mapping, as a method to verify the area and classification accuracy of the image in order to obtain the accuracy and reliability of identification and evaluation of the Shade shed area for P. notoginseng. Result The accuracy of decision tree classification method is 87%, and that of visual interpretation method is 99%. The area accuracy of decision tree classification can reach 80%. Conclusion Compared with the traditional method of using high-resolution commercial images and artificial drawing spots, this method of extracting the P. notoginseng area could quickly determine P. notoginseng planting resources with higher accuracy, thereby providing basic data and effective technical methods for scientific supervision and value estimation of locally cultivated agricultural products.

关 键 词：森林测计学 种植信息提取 高分1号 决策树 精度分析 三七 

分 类 号：S758.5[农业科学—森林经理学]