高光谱遥感数据用于水稻精细分类研究  

Airborne Hyperspectral Remote Sensing Image Data is Used for Rice Precise Classification

作　　者：张丰[1] 熊桢[2] 寇宁[1] 

机构地区：[1]武汉大学 [2]中科院遥感所

出　　处：《武汉理工大学学报》2002年第10期36-39,共4页Journal of Wuhan University of Technology

基　　金：中科院"九五"重大资助项目 (KZ95 1- A1- 30 2 - 0 4 - 0 1);中国资源环境遥感信息系统及农情速报资助项目 (KZ95 T- 0 3- 0 4 -0 1)

摘　　要：根据水稻生长期的高光谱数据的光谱特征 ,设计了一个混合决策树分类算法。该算法的特征波段根据波段间的可分离性决定 ,算法的选择根据实际分类效果决定 ,波段间的可分离性根据各波段均值的归一化距离决定。最后用江苏常州市金坛良种场的高光谱图象数据做了分类实验 ,取得了测试样本总体分类精度 94 .According to the rice spectral features of hyperspectral image data acquired during the rice is growing, a hybrid decision tree classification algorithm dealing with the variety of rice is developed. The feature bands are selected according to the separability among bands, but the classification algorithm is selected according to its classification result. The separability among bands is calculated by the normalized mean value of each band. In the end, a classification experiment is done. In the experiment the hyperspectral image data acquired in Jintan rice breeding farm is used. A good classification result is achieved. The classification accuracy of test samples is reached 94.9 percent.

关 键 词：高光谱遥感数据 混合决策树 水稻 精细分类 

分 类 号：S511[农业科学—作物学]