川中丘陵地区撂荒地的遥感提取方法研究

Research on remote sensing extraction methods for abandoned land in hilly areas of central Sichuan

四川省以丘陵地貌为主,地块破碎耕作条件较差,近年耕地撂荒问题日益严峻。目前采用遥感方法提取撂荒地是主要手段之一,采用遥感方法高精度提取丘陵地区撂荒耕地的研究甚少。文章以资阳市祥符镇为例,采用哨兵2号和谷歌高清卫星遥感影像为数据源结合纹理以及光谱特征对研究区2021—2022年度耕地撂荒情况进行监测,结合Modis年际NDVI变换情况采用NDVI阈值法辅以提取精度校正,结果表明提取精度高达92%。

Sichuan Province is dominated by a hilly landform,with poor farming conditions.In recent years,the problem of abandoned cultivated land has become increasingly serious.At present,the remote sensing method is one of the main methods to extract abandoned farmland,and there is little research on high-precision extraction of abandoned farmland in hilly areas by the remote sensing method.Taking Xiangfu Town,Ziyang as an example,this paper uses Sentinel 2 and Google HD satellite remote sensing images as data sources,combined with texture and spectral characteristics,to monitor the abandoned cultivated land in the study area from 2021 to 2022,and adopts the NDVI threshold method combined with extraction accuracy correction combined with Modis interannual NDVI transformation.The results show that the extraction accuracy is as high as 92%.