基于高分二号影像的森林和耕地的提取方法

Extraction Method of Forest and Cultivated Land Based on GF-2 Data

针对在高分辨率遥感图像中提取森林和耕地时,湖泊在不同光照条件下的颜色变化、山体阴影等情况导致的森林和耕地提取中出现错分,识别准确率低的问题,提出了一种结合归一化水体指数法与最大似然分类方法进行遥感图像森林和耕地的提取,通过实验表明该方法能够达到较高的识别效果,分类准确率达85%,鲁棒性良好。

Aiming at the problems of high-resolution remote sensing images in the extraction of forest areas and farmland,the color change of lakes under different light conditions,and other errors in forest extraction and farmland cause low recognition accuracy,this paper proposes a method of combining normalized water bodies and maximum likelihood classification method to be used to extract the forest of remote sensing images.Experiments show that the method can achieve higher recognition effect,the classification accuracy rate is 85%,and the robustness is good.