“撂荒地”影像特征分析及信息提取——以右玉县牛心堡乡为例  

“Abandoned Farmlands”image features analysis and information extraction——In case of Niuxinbao township in Youyu

作　　者：王利[1] 李欣欣 

机构地区：[1]辽宁师范大学城市与环境学院,辽宁大连116029

出　　处：《辽宁师范大学学报（自然科学版）》2017年第4期519-524,共6页Journal of Liaoning Normal University:Natural Science Edition

基　　金：国家科技支撑计划资助项目(2008BAH31B06);部省共建人文社会科学重点研究基地资助项目(14JJD790038)

摘　　要：耕地撂荒严重影响了我国粮食产量,成为近20年来耕地利用中出现的重大问题.多光谱遥感具有数据获取范围广、速度快等特点,在国土资源研究中具有相当大的潜力和优势.为提取和识别撂荒地,研究采用2010—2014年各年份的4—10月多光谱遥感数据Modis/Terra的NDVI数据产品和2014年6月份Landsat 8遥感影像数据,生成撂荒地、林地、耕地类型采样点的NDVI时间序列曲线,利用时间序列曲线来表征撂荒地、林地及耕地的生命周期特征,结合3种覆被类型的生命周期特征及其NDVI曲线特征,对3种覆被NDVI曲线变化趋势进行对比分析,从而区分撂荒地、林地、耕地,以达到识别撂荒地的目的.结合Landsat 8遥感影像判读和野外考察验证样本,最后对遥感影像进行判读,分析撂荒地的光谱、纹理、形状及位置特征,确立撂荒地解译标志,运用ArcGIS 10.2平台实现撂荒地信息提取.结果表明,运用多光谱遥感技术能够很好地提取撂荒地,提取撂荒地精度达到86%.因此,多光谱遥感技术在撂荒地识别研究中具有相当的可行性.The abandonment of arable land has seriously affected the grain output in our country and has become a major problem in the utilization of cultivated land in the past two decades.In this paper,the NDVI data from Modis/Terra and the Landsat 8 remote sensing image data of June,2014,from April to October in 2014,and the NDVI time series of abandoned land,forest land and cultivated land are characterized by time series NDVI.The NDVI time series curves of abandoned land,forest land and cultivated land type are generated.Combined with the growth characteristics and NDVI curve characteristics,three NDVI curves were analyzed to identify abandoned land.Combined with Landsat 8 remote sensing image and field survey data to establish the interpretation of signs we made an analysis of the abandoned land spectrum,texture,shape and location characteristics,and also use the ArcGIS to achieve the extraction of abandoned land information.

关 键 词：MODIS数据 植被指数 撂荒地 牛心堡乡 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]