多光谱图像中云层及阴影的检测与消除

云层及阴影的检测是遥感应用领域的一个非常重要的问题。目前这方面的资料比较少，尤其是在被动接收条件下。提出了一种云层及阴影检测的算法，该算法的主要依据是云层及阴影覆盖区域高平滑性的特征，算法主要包括图像变换、亮度校正、云层及阴影的检测、滤波四个步骤。相对于现存方法而言，优点在于整个算法适合主被动接收条件，算法的处理比较简单，算法实现过程中没有很苛刻的参数选择，算法的适用范围较广。仿真结果不仅验证了上面的分析，而且表明该算法去云层及阴影的效果比较理想。

Detection and removal of clouds and their shadows from multi-spectral image

It is an important problem to detection of clouds and their shadows for remote sensing application.Relatively little work had appeared in present literature in this area,especially on the condition of passive receiver.In this paper,a new scheme to detect clouds and their shadows was purposed mainly based on feature that their brightness changes corresponding to areas of clouds and their shadows appear much smoother than uncontaminated areas.The algorithm mainly includes four steps:image transform,correction of brightness,detection of clouds and their shadows,filtering.Compared with present method,the algorithm is adapted to condition of both active and passive accepter,and simple to realize.During the implementation of the algorithm,there was not demanding parameters to be selected,so the algorithm is also adapted widely to other kinds of remote sensing image.The simulation results has testified above analysis and show that the algorithm effectively detects and removes clouds and their shadows.