资源三号卫星在种植型药用植物资源调查中应用研究-以吉木萨尔县红花为例.pdfVIP

资源三号卫星在种植型药用植物资源调查中的应用研究 —以吉木萨尔县红花为例 1 3 1，2* 3 1 1 森巴提 ，贾盛杰 ，郑江华 ，石明辉 ，娜仁花 ，郑淑丹 ，郭宝 3 1，2 3 林 ，孙志群 ，李晓瑾 1、新疆大学资源与环境科学学院，新疆 乌鲁木齐 830046；zheng_jianghua@126.com 2 、新疆大学智慧城市与环境建模普通高校重点实验室，新疆 乌鲁木齐 830046； 3 、新疆维吾尔自治区中药民族药研究所，新疆，乌鲁木齐 830002 摘要：【目的】本文以红花为研究对象，以遥感技术为基础，探索药用植物资源调查的方法。【方法】以红 花的一个典型分布区新疆昌吉州吉木萨尔县作为研究样区，对研究样区开展野外遥感调查获取红花的生境 及判读特征；首次采用中国资源三号卫星（ZY-3 ）影像对红花监测并进行图像预处理，完成基于主成分分 析和纹理特征分析的最大似然监督分类法的分类，计算分类结果，得出分类精度。【结果】其样本准确率为 68.75%，总体分类精度达到 68.757%，卡帕系数为 0.573，分类结果为中度一致性。【结论】可知利用遥感 技术解译药用植物有一定的难度，结果表明利用中国资源三号卫星对药用植物红花的遥感资源调查基本是 可行的。 关键词：资源三号卫星；遥感技术；药用植物资源调查；红花 Abstract ：[Objective] In this paper, we took Carthamus tinctorius L as the research object to explore resources of medicinal plants in China based on remote sensing technology. [Method] We selected a typical region, Jimsar County of Changji Hui Autonomous Prefecture in Xinjiang, as the sample plot. On this area, we got some features of habitat of Carthamus tinctorius L through filed survey. ZY-3 satellite images had been chosen to finish our interpretation work. Finally, we preprocessed these images and produced the classified figures by the method of maximum likelihood supervised classification based on principal component analysis and texture features. [Result] The results showed that sample accuracy rate was 68.75%, the overall classification accuracy was up to 68.757% and the Kappa coefficient was 0.573. The classification result was moderate consistency. [Conclusion] Generally, we could infer that it is feasible to use ZY-3 satellite to explore resources of medicinal plants, such as Carthamus tinctorius L. However, there are some difficulties in interpreting medicinal plants by using remote sensi