摘要
遥感分类技术是获取土地利用/覆盖数据的主要方法。分层分类思想强调将分类过程逐级进行,每层选用不同的分类标准和方法;监督分类是基于传统统计分析的分类法,具有算法成熟,简便易行的特点。将2种方法相结合,建立起一个复合分类模型,并在SPOT影像上进行试验。试验证明:该方法能有效地提高分类精度,比单一使用监督分类法得到的结果精度提高了8.41%。
The technology of remote sensing image classification is the main means of LUCC research. The stratified classification is a method based on the idea of division of layers step by step and different criteria and methods in each layer; while the supervised classification highlights traditional statistical and analytic approach, which is a mature algorithm, simple and convenient as well as easily being operated. Combining these two methods is to establish a compound classification model which is tested in SPOT im-age. The result shows that this method can effectively improve precision of classification. In detail, com-pared with supervised classification method, the result from compound method improves 8.41% of the precision.
出处
《林业调查规划》
2007年第4期37-39,44,共4页
Forest Inventory and Planning
基金
吉林省科技厅重点项目
中国科学院遥感信息科学重点实验室开放基金支持
关键词
分层分类
监督分类
遥感分类
决策树分类器
复合分类模型
遥感影像
stratified classification
supervised classification
remote sensing classification
classifica-tion frame of decision tree
compound model
remote sensing image
