摘要
高原山区多是山地、丘陵,地形崎岖,农户种植的耕地破碎、作物种类繁多是造成遥感分类结果较差的主要因素。针对这个问题,以云计算引擎(GEE)为数据处理平台,利用集成学习方法对高原山区水稻进行遥感分类。实验结果表明集成学习后的强分类器比单一分类器的泛化性能强,分类的总体精度86.61%和Kappa系数0.837 8,水稻的用户精度84.10%和制图精度75.51%,说明集成学习方法对地块破碎、土地利用类型复杂的高原山区水稻信息提取具有一定的参考意义。
Most of the plateau mountains are mountains and hills,the terrain is rugged,the broken cultivated land planted by farmers and a wide variety of crops are the main factors leading to the poor results of remote sensing classification.In order to solve this problem,using the cloud computing engine(GEE) as the data processing platform,the ensemble learning method is used for remote sensing classification of rice in plateau and mountain areas.The experimental results show that:the generalization of the strong classifier after ensemble learning is stronger than that of a single classifier,the overall accuracy of classification is 86.61% and the Kappa coefficient is 0.837 8,the user accuracy of rice is 84.10% and the producer accuracy is 75.51%,indicating that the integrated learning method has a certain reference significance for rice information extraction in plateau and mountainous areas where the land is broken and the land use type is complex.
作者
陈波
王加胜
王刚
苗旺元
Chen Bo;Wang Jiasheng;Wang Gang;Miao Wangyuan(School of Information Science and Technology,Yunnan Normal University,Kunming 650500,Yunnan,China;The Engineering Research Center of GIS Technology in Western China,Ministry of Education of China,Kunming 650500,Yunnan,China)
出处
《计算机应用与软件》
北大核心
2023年第10期242-249,共8页
Computer Applications and Software
基金
国家自然科学基金项目(41961056)。