中文摘要：

为探讨利用国产HJ卫星遥感数据进行农作物种植面积估算及农业气象灾害定量监测评估的可行性，以江汉平原南部县市为研究区，利用2011年6景不同时相的HJ卫星CCD遥感数据，结合野外调查样方点和面积较大的试验基地样本，通过分析研究区主要秋收作物（棉花、一季中稻）不同生育期的光谱特征和归一化植被指数（NDVI）时序变化特征，对分类影像进行序列阈值分割、掩膜处理，最后利用决策树算法成功提取了棉花和一季中稻的种植面积，得到面积精度和样本点精度均大于90％，Kappa系数为0．9835的结果。只要不受到云和降水的影响，能获取江汉平原区域关键时相（5月中下旬、6月上旬、7月上旬、7月下旬或8月上中旬、10月上旬）HJ卫星CCD影像数据，便能很好地应用于江汉平原棉花和中稻作物提取。

英文摘要：

This study is to examine the feasibility of using Chinese HJ-CCD remote sensing data to estimate cotton and mid-season rice planting area in southern Jianghan Plain, Hubei Province. Filed investigation has been done for the study. Spectral property and NDVI index during the growing period analyzed on the six available HJ-CCD images in 2011. Threshold configurations were set to build decision trees for cotton and mid-season rice planting area estimation. Finally, planting area of cotton and mid-season rice were extracted with decision trees method. The precision testing result show that area accuracy and sample point accuracy were not less than 90% and Kappa coefficient was 0. 983 5. Therefore, it is concluded that cotton and mid-season rice planting area could be estimated precisely as long as you can get the key HJ Satellite CCD Data during cloudless days in late May, early June, early July, late July or early August or middle of August and early October. It should be able to provide scientific basis for planting structure adjustment and evaluation work of agro-meteorological disaster.