海岛区域含云雾影像预处理技术研究

薄雾、朵云的影响导致光学遥感影像质量下降,欲有效开展海岛区域测绘需对相关影像进行必要的预处理。首先分析了海岛区域影像中薄雾和朵云的辐射及几何特点,其次采用改进的基于暗原色先验知识的薄雾处理方法和朵云及云影双阈值同时检测方法T多源影像配准后压盖地物信息替换,以及改进的影像匀光匀色方法。实验证明,上述方法能够有效地改善海岛区域影像的视觉质量,提高其作业效率。

Study on Dehazing and Cirrus Processing of Island Area Image

Due to the poor-quality optical remote sensing image caused by haze and cirrus,preprocessing is necessary before surveying and mapping launches in island area image. Firstly, radiant and geometrical characteristics of haze and cirrus are analyzed in island area image. Secondly, several approaches has been conducted to process the image,they include: the improved algorithm of dehazing based on dark channel prior; the method of cirrus and their cast shadows detection based on double thresholds; the substitute of covered ground features after registration of multi-temporal image data and the improved method of image dodging. The experiments show that the above approaches have effectively improved the visual quality in island area image, and the operation efficiency will be raised.