基于光谱特征分异的玉米种植面积提取
发布时间：2018-07-23 20:18
【摘要】：玉米种植面积的准确获取是进行玉米长势监测和产量估测的前提与基础。在对Landsat-8/OLI影像进行辐射定标、大气校正、几何精校正和裁剪等预处理的基础上,基于典型地物光谱空间差异与物候特征的异同,选取具有代表性的4种植被指数[归一化差值植被指数(NDVI)、差值植被指数(DVI)、比值植被指数(RVI)、绿度植被指数(GVI)]和近红外波段反射率,通过构建植被光谱特征指标阈值对不同地物进行识别和分类,最后获取玉米种植面积。结果表明,利用近红外波段反射率可以将农作物与其他地物区分开来,即当其反射率值大于0.37时,地物为农作物。对不同种类农作物识别时,选择NDVI0.86、DVI0.53、RVI13.00、GVI3 713.60作为分类阈值,可以将玉米与水稻和大豆区分,准确提取到玉米的种植面积。利用样本数据和当地农业部门提供的数据进行面积提取精度验证,总体精度为92.75%,说明基于多光谱特征指标建立分类阈值的方法可以准确提取玉米种植面积,该方法可以为江淮玉米种植区县域玉米种植面积的提取提供参考。
[Abstract]:Accurate acquisition of maize planting area is the precondition and basis for maize growth monitoring and yield estimation. On the basis of radiometric calibration, atmospheric correction, geometric precision correction and trimming of Landsat-8/OLI images, based on the similarities and differences between the spatial differences of typical geographical features and phenological characteristics, the representative 4 cropping fingers are selected. The number [normalized difference vegetation index (NDVI), differential vegetation index (DVI), ratio vegetation index (RVI), green vegetation index (GVI) and near-infrared reflectance were identified and classified by constructing the threshold value of vegetation spectral characteristic index, and the maize planting area was obtained at last. The results showed that the reflectance of near infrared band could be used. To distinguish the crops from other objects, that is, when their reflectivity is greater than 0.37, the land is a crop. When identifying different kinds of crops, NDVI0.86, DVI0.53, RVI13.00, and GVI3 713.60 are selected as the classification threshold, and corn and rice and soybean can be distinguished, and the planting area of Maize can be extracted accurately. The data provided by the industry department verified the accuracy of area extraction, and the overall precision was 92.75%. It indicated that the method of setting up the classification threshold based on the multi spectral characteristic index could accurately extract the maize planting area. This method could provide reference for the extraction of maize planting area in the county area of the Jianghuai corn planting area.
【作者单位】： 安徽农业大学资源与环境学院;江苏省农业科学院农业信息研究所;国家农业信息化工程技术研究中心;
【基金】：国家自然科学基金项目(41571323) 江苏省重点研究计划项目(BE2016730)
【分类号】：S127;S513
本文编号：2140529
[Abstract]:Accurate acquisition of maize planting area is the precondition and basis for maize growth monitoring and yield estimation. On the basis of radiometric calibration, atmospheric correction, geometric precision correction and trimming of Landsat-8/OLI images, based on the similarities and differences between the spatial differences of typical geographical features and phenological characteristics, the representative 4 cropping fingers are selected. The number [normalized difference vegetation index (NDVI), differential vegetation index (DVI), ratio vegetation index (RVI), green vegetation index (GVI) and near-infrared reflectance were identified and classified by constructing the threshold value of vegetation spectral characteristic index, and the maize planting area was obtained at last. The results showed that the reflectance of near infrared band could be used. To distinguish the crops from other objects, that is, when their reflectivity is greater than 0.37, the land is a crop. When identifying different kinds of crops, NDVI0.86, DVI0.53, RVI13.00, and GVI3 713.60 are selected as the classification threshold, and corn and rice and soybean can be distinguished, and the planting area of Maize can be extracted accurately. The data provided by the industry department verified the accuracy of area extraction, and the overall precision was 92.75%. It indicated that the method of setting up the classification threshold based on the multi spectral characteristic index could accurately extract the maize planting area. This method could provide reference for the extraction of maize planting area in the county area of the Jianghuai corn planting area.
【作者单位】： 安徽农业大学资源与环境学院;江苏省农业科学院农业信息研究所;国家农业信息化工程技术研究中心;
【基金】：国家自然科学基金项目(41571323) 江苏省重点研究计划项目(BE2016730)
【分类号】：S127;S513
【相似文献】
相关期刊论文 前5条
1 吴豪翔,王人潮;土壤光谱特征及其定量分析在土壤分类上的应用研究[J];土壤学报;1991年02期
2 张艳彬;王玉;杨忠芳;陈岳龙;;成都经济区土壤光谱特征及其意义[J];世界地质;2007年01期
3 罗清;;青藏高原季节湖土壤光谱特征研究[J];西南大学学报(自然科学版);2012年07期
4 张琳,刘培君,常萍,艾力西尔·库尔班;阜康试验区主要地物光谱特征及其分析[J];新疆气象;1995年05期
5 韩阳;赵云升;王野乔;;不同含水量土壤偏振光谱特征定量分析[J];遥感学报;2013年05期
相关硕士学位论文 前2条
1 刘效栋;黄土台塬区土壤有机质高光谱特征及反演研究[D];西北农林科技大学;2015年
2 王祥峰;基于光谱特征以及养分指示因子的土壤养分遥感监测研究[D];辽宁工程技术大学;2015年
，本文编号：2140529
本文链接：https://www.wllwen.com/kejilunwen/nykj/2140529.html
