RainCityscapes

该数据集名为RainCityscapes，包含了使用深度信息合成的雨景图像，这些图像模拟了不同强度的雨痕和雾气，并按照雾气浓度分为三个等级。此外，该数据集提供了9432张用于训练的雨景图像和1188张用于测试的雨景图像，以及相应的无雨背景图像。总规模达到了10620张雨景图像。该数据集的任务是图像去雨。

This dataset, named RainCityscapes, consists of rainy scene images synthesized using depth information. These images simulate rain streaks and fog with varying intensities, and are categorized into three levels based on fog concentration. Additionally, it provides 9432 rainy images for training, 1188 rainy images for testing, along with their corresponding rain-free background images. The total number of rainy scene images in this dataset reaches 10620. The task of this dataset is image deraining.