多时相高分一号影像在丘陵地区大宗农作物提取中的应用

发布时间：2018-03-23 03:05

本文选题：多时相　切入点：高分一号(GF-)　出处：《南方农业学报》2017年01期 　论文类型：期刊论文

【摘要】：【目的】基于多时相的高分一号(GF-1)影像,利用面向地块对象分类法提取广西崇左市江州区大宗农作物种植面积,为南方多云雨丘陵地区提取作物信息提供参考。【方法】以2 m分辨率的GF-1影像为数据源,采用人机交互的方式准确识别地表覆盖的地块信息,基于对多时相GF-1影像进行云影检测,并处理生成影像的光谱、归一化植被指数(NDVI)、亮度等特征,采用面向地块对象的分类方法提取甘蔗、水稻和香蕉的作物信息。【结果】根据混淆矩阵评价分类的结果可知,提取大宗农作物的总体精度为90.08%,Kappa系数达0.85,满足农业成果应用的精度要求。【结论】利用有效影像数据,结合地块数据完成作物信息提取,该技术方法能够准确提取丘陵地区大宗农作物信息,为解决南方多云雨丘陵地区提取作物信息难题提供了有效途径。
[Abstract]:[objective] based on the multitemporal GF-1 image, the large crop planting area in Jiangzhou area, Chongzuo City, Guangxi was extracted by using the massive-oriented object classification method. This paper provides a reference for the extraction of crop information in the southern cloudy and rainy hilly areas. [methods] based on 2 m resolution GF-1 image as data source, human-computer interaction is used to accurately identify the land cover information. Based on the cloud image detection of multitemporal GF-1 image and the spectral processing of the generated image, normalized vegetation index (NDVI), brightness and other features, sugarcane was extracted by using the classification method oriented to field objects. [results] according to the results of the confusion matrix evaluation and classification, the total precision of extracting large crops was 90.08 and the Kappa coefficient was 0.85, which met the precision requirements of agricultural application. [conclusion] effective image data were used. The method can extract crop information accurately in hilly area, and it provides an effective way to solve the problem of crop information extraction in the southern cloudy and rainy hilly area.
【作者单位】：广西农业科学院农业科技信息研究所;
【基金】：广西科学研究与技术开发计划项目(桂科合14125008-1-6) 广西农业科学院科技发展基金项目(2015JZ140)
【分类号】：S127

【相似文献】