摘要
遥感图像快速识别是遥感数据应用于土地利用/土地覆盖动态监测的难点问题。根据海拔高差大地区地表覆盖的垂直性特点,利用数字高程模型分割影像,分析各区域的光谱特征和纹理特征,采用植被指数和纹理组合对不同区域地表信息进行提取,并根据野外实测数据进行精度验证。结果表明,该方法能有效提高地物识别精度,有一定实用价值。
Remote sensing image analysis on land use/land cover （LUCC） information is an important preparation work for the macroscopical geoscience research.Based on the verticality characters of the vegetation distribution,the digital elevation model （DEM） is established,and using it,the whole image is divided into several parts.Then the surface cover vegetation index and the texture feature are analyzed respectively,and using different wave bands combinations,certain operations are made to gain the useful information.The final precision analysis indicates that this method is suits to the test region,and it can be applied in the adjacent area.
出处
《科学技术与工程》
北大核心
2014年第36期141-145,共5页
Science Technology and Engineering
基金
中央高校基本科研业务费专项资金(CUGL120261)资助
关键词
分层
数字高程模型
植被指数
纹理
LANDSAT
multi-layer
digital elevation model （DEM）
vegetation index
texture feature
landsat
