摘要
红枣是新疆特色林果业的重要组成部分,为快速准确地获取红枣的种植信息,适应农业管理现信息化,该文以新疆阿拉尔地区为例利用Landsat8遥感卫星影像结合GoogleEarth以及户外实地定标,通过训练样本进行机器学习后进行支持向量机分类将红枣分类,并通过Majority和Minority分析与聚类分析分别对分类结果进行优化,结果表明:两种分类优化后的准确度分别为93.7%与89.5%以上。该方法运用遥感技术获取计算枣树种植信息,可为相关管理部门在红枣种植面积的估算中提供一个实时有效的参考方法。
Jujube is an important part of Xinjiang's characteristic forest and fruit industry.In order to quickly and accurately obtain planting information of jujube,and to adapt to the current informatization of agricultural management,this article uses the Landsat_8 remote sensing satellite image combined with GoogleEarth and outdoor field calibration in the Alar region of Xinjiang as an example.The training samples are machine-learned and then classified by support vector machine.The dates are classified by Majority and Minority analysis and cluster analysis.The results show that the accuracy of the two classifications is 93.7%and 89.5%,respectively.the above.This method uses remote sensing technology to obtain information about jujube planting,which can provide a real-time and effective reference method for the relevant management departments in the estimation of jujube planting area.
出处
《科技资讯》
2020年第14期56-61,共6页
Science & Technology Information
基金
塔里木大学现代农业工程重点实验室开放课题“基于3S技术的阿克苏地区红枣种植信息提取及系统开发”(项目编号:TDNG20150502)。
