作者:Kun JIA,Jingcan LIU,Yixuan TU,Qiangzi LI,Zhiwei SUN,Xiangqin WEI,Yunjun YAO,Xiaotong ZHANG
摘要:The newly launched GF-2 satellite is now the most advanced civil satellite in China to collect high spatial resolution remote sensing data.This study investigated the capability and strategy of GF?2 multispectral data for land use and land cover (LULC) classification in a region of the North China Plain.The pixel-based and object-based classifications using maximum likelihood (MLC) and support vector machine (SVM) classifiers were evaluated to determine the classification strategy that was suitable for GF?2 multispectral data.The validation results indicated that GF-2 multispectral data achieved satisfactory LULC classification performance,and object-based classification using the SVM classifier achieved the best classification accuracy with an overall classification accuracy of 94.33% and kappa coefficient of 0.911.Therefore,considering the LULC classification performance and data characteristics,GF-2 satellite data could serve as a valuable and reliable high-resolution data source for land surface monitoring.Future works should focus on improving LULC classification accuracy by exploring more classification features and exploring the potential applications of GF-2 data in related applications.
发文机构:State Key Laboratory of Remote Sensing Science Beijing Engineering Research Center for Global Land Remote Sensing Products Institute of Remote Sensing and Digital Earth Beijing Geoway Times Software Technology Co.
关键词:LANDuseandLANDCOVERCLASSIFICATIONGF-2NORTHChinaPLAINMULTISPECTRALdata
分类号: P[天文地球]