The aim of this paper is to improve the accuracy of classification of crops plant distribution using sentinel ASAR (Advanced Synthetic Aperture Radar) data. In this paper, the study case place is Jianli county, Hubei Province,which is the largest rice planting area in China, the main method is following, extracting the crop field object border using GF-1 satellite multi-bands optical data based on image segment and merge technical method; averaging the radar backscatter coefficient of ASAR data within each object,this can eliminate the influence of coherent speckle noise; Through the analysis of the ASAR backscattering characteristics of early season rice paddy, found that the backscattering coefficient value of early season rice paddy is higher in Pre-irrigation stage, lower in irrigation stage and raising in seedling stage,built the early season rice pad-dy identification formula according these rules. Extraction the early season rice planting distribution information in 2015 at Jianli county. According to compared with investigating in place with 23.4 km2 area, the Kappa coefficient is 0.83, The accu-racy is improved by only using the ASAR data extraction method. This method is inexactitude to optical remote sensing satel-lite acquisition date, and ASAR data gained by sentinel 1A satellite is not affected by the clouds, can be obtained regular-ly, so this method is suitable for crop cultivation spatial distribution information extraction of business operation.%针对目前利用ASAR数据提取作物种植空间分布的精度因相干斑点噪声问题而达不到业务运行要求问题,以全国水稻种植面积最大的湖北省监利县为例,运用空间分辨率为16 m的高分一号多光谱数据,采用图像切割和融合方法提取农田边界,将早稻早期三景ASAR数据以每个对象(一个对象代表一块农田)取均值,以此消除相干斑点噪声,结合早稻早期ASAR数据变化特征(早稻灌水前期ASAR后向散射系数大,灌水后最小,秧苗后期变大),提取监利县2015年早稻种植空间分布.通过与调查区(面积23.4 km2)对比,其Kappa系数为0.83,其精确度较单独用ASAR数据提取方法提高一个层次,该早稻提取方法适合业务化运行.
展开▼
