数学形态学在遥感影像云层识别中的应用

Cloud determination in remote sensing images based on mathematical morphology

针对传统遥感影像质量评价中云层覆盖量无法定量评价的问题,文章提出采用利用边缘信息的阈值分割结合数学形态学方法来提取遥感影像云层覆盖范围。利用边缘信息的阈值分割方法能够有效利用影像自身信息来改善分割结果,再结合形态学方法,进而能消除道路、房屋等大部分噪声信息,最终实现遥感影像上不同特征云层覆盖范围的自动提取。基于浙江全省高分辨率遥感影像的实验结果表明:该方法能够快速有效地识别出遥感影像上云层覆盖范围,研究结果对于遥感影像云层覆盖的自动评价具有参考价值。

To solve the quality evaluation for the cloud cover of the high-resolution remote sensing Im- ages, this paper proposed an extracting cloud method based on mathematical morphology and used edge data to improve the global thresholds, in order to effectively use the edges between objects and back- ground and improve image segmentation results. Most noise such as buildings could be removed by using mathematical morphology method. Experimental result indicated that the method could effectively extract the cloud cover from high-resolution remote sensing images.