基于GF-1 WFV数据的甘蔗种植面积监测

Monitoring of planting area of sugarcane based on GF-1 WFV data

基于高分一号卫星(GF-1)的宽幅相机(WFV)数据,以广西扶绥县为例,引入坡度、耕地基础地理数据库。根据遥感影像的光谱特征,采用非监督分类、监督分类及NDVI剔除法相结合的方法,准确提取出该县甘蔗种植面积。经精度验证,GF-1WFV数据对扶绥县种植面积遥感监测总体精度为总精度达89.15%,Kappa系数为0.7803,生产者精度和用户精度分别为90.72%和87.21%,均达到农业生产的需求。结果表明,GF-1 WFV数据能为县域尺度甘蔗种植面积的进行有效的追踪和检测。

Based on GF-1 WFV data from the wild field camera（WFV）of GF-1 satellite of China, the basic geographic database of slope and farmland in Fusui County of Guangxi was introduced. According to the spectral characteristics of remote sensing image, the sugarcane planting area in this county was extracted accurately by the combination use of unsupervised classification, supervised classification and NDVI culling method. The total precision of remote sensing monitoring of planting area in Fusui County by GF-1 WFV data is 89.15%, the kappa coefficient is 0.7803. The producer accuracy and user accuracy are 90.72% and 87.21 respectively, which all meet the requirements of agricultural production. The results show that GF-1 WFV data can be used to track and detect the planting area of sugarcane in county scale.