摘要
针对纹理特征窗口大小的选择对SAR图像分类精度有很大影响的情况 ,在传统的最大似然分类法(ML)的基础上 ,提出了一种将大小窗口结合在一起的改进纹理特征分类方法 ,并分别用该方法与传统的ML法对哈尔滨附近地区 5 12× 5 12大小的JERS - 1SAR图进行分类 。
In view of the situation that the choice of window size of texture characteristics has a great effect on the accurate rate of SAR image classification, on the base of the traditional maximum likelihood classification method, an improved classification method by use of texture characteristics with combination of large and small windows is proposed in this paper. The improved method and the traditional one are respectively used to classify the 512×512 JERS-1SAR image near Harbin, and the experimental results show the feasibility and effectiveness of the new method.
出处
《系统工程与电子技术》
EI
CSCD
2000年第4期15-17,共3页
Systems Engineering and Electronics
基金
国防预研基金资助课题!(Y96 - 0 1 )
关键词
合成孔径雷达
图像处理
遥感成像
图像纹理
SAR Image processing Remote imaging Maximum likelihood classification