基于HJ-1B卫星遥感数据的水稻识别技术研究

为快速、准确地在遥感图像上识别水稻作物的信息,满足县级尺度水稻遥感监测的需要,以野外实地调查资料、1∶5万地形图数据为辅助,通过光谱分析法,分析研究HJ-1B星CCD数据的水稻作物的光谱反射特性,建立水稻作物遥感信息识别模型。采用决策树分类方法提取水稻作物信息,并将该技术方法应用于广西宾阳县水稻作物信息提取研究。采用实测样地数据,利用混淆矩阵进行精度评价验证,总精度为94.9%,Kappa系数为0.8533。研究表明,该水稻作物的识别技术,可以为了解我国水稻种植情况,进行水稻长势监测和产量估测提供技术参考。

Study on Rice Identification Technology Using HJ-1B Data

To identify and extract rice information from remote sensing images rapidly,correctly and objectively,satisfied the needs of rice remote sensing monitoring in a large scale,the spectral reflection characteristic of rice was analyzed by spectral analysis method,based on HJ-1B data.With field survey data,1∶5 million digital elevation model (DEM) data for the auxiliary parameters,a remote sensing information identification model for rice was established.Using decision tree classification method to extract the information of rice.And the rice information was extracted in Binyang country,Guangxi province by using this method.The accuracy of this rice identification model was verified by GPS measured sample data and the precision reached 94.9%,Kappa is 0.8533.Results show that this rice identification technology can provide a technological support for understanding rice planting situation,monitoring rice growing condition and estimating the yield of rice.