单 捷,邱 琳,孙 玲,王志明.基于Radarsat-2的冬小麦种植面积提取方法研究[J].麦类作物学报,2017,(9):1209 |
基于Radarsat-2的冬小麦种植面积提取方法研究 |
Extraction Method of Winter Wheat Planting Area Based on Radarsat-2 Data |
|
DOI:10.7606/j.issn.1009-1041.2017.09.11 |
中文关键词: Radarsat-2 遥感 冬小麦 种植面积 提取方法 |
英文关键词:Radarsat-2 Remote sensing Winter wheat Planting area Extraction method |
基金项目:国家科技重大专项课题(09-Y30B03-9001-13/15-4);江苏省农业科学院基金项目(6111651,6111650);农业部遥感应用中心技术创新课题(2911660);江苏省农业科学院基本科研业务专项(ZX-15-3003);江苏省农业科技自主创新资金项目[CX(17)3020] |
|
摘要点击次数: 1156 |
全文下载次数: 577 |
中文摘要: |
为探寻基于Radarsat-2的冬小麦种植面积提取方法,以设立在江苏盐城研究区的冬小麦为研究对象,选用2014年3月3日-2014年6月7日期间5期Radarsat-2全极化影像,采用支持向量机法和最大似然法分别对各时相的冬小麦种植面积进行提取,并以地面实测GPS样方进行精度验证。结果表明,以支持向量机法和最大似然法提取冬小麦面积的精度均在4月20日达到最高,分别为66.4%和63.9%。对4月20日支持向量机法的冬小麦面积提取结果进一步进行耕地地块优化和碎小图斑去除处理后,冬小麦面积的提取精度可提高到79.6%。 |
英文摘要: |
In order to study the extraction method of winter wheat planting area based on Radarsat-2 data,five scenes of Radarsat-2 satellite images collected from March 3, 2014 to June 7, 2014 were used to extract winter wheat planting area through the methods of Support Vector Machine(SVM) and Maximum Likelihood Classification(MLC), and the accuracy of this investigation is verified by on-site GPS measurement quadrat areas. The results indicated that the highest extraction accuracies of both methods were achieved on April 20, so we chose the image of April 20 to study the effect of farmland parcel optimization and classification patch optimization on SVM's accuracy. The extraction accuracy of SVM was improved from 66.4% to 79.6% after optimization. |
查看全文 查看/发表评论 下载PDF阅读器 |
关闭 |