Clearing the Skies: A Deep Network Architecture for Single-Image Rain Removal
Clearing the Skies: A Deep Network Architecture for Single-Image Rain Removal
复制标题
清除天空：用于单图像除雨的深层网络架构
DOI：
10.1109/tip.2017.2691802
复制
发表时间：
2017-06-01
影响因子：
10.6
通讯作者：
Paisley, John
中科院分区：
文献类型：
--
作者：
Fu, Xueyang;Huang, Jiabin;Paisley, John
我们引入了一种名为DerainNet的深度网络架构，用于从图像中去除雨纹。基于深度卷积神经网络（CNN），我们直接从数据中学习雨天和干净图像细节层之间的映射关系。因为我们不具备与真实世界的雨图像相对应的地面真实值，所以我们将雨图像合成用于训练。与其他增加网络深度或广度的常见策略相比，我们使用图像处理领域知识来修改目标函数，并使用适度大小的CNN来改进去训练。具体来说，我们在细节（高通）层而不是图像域中训练我们的DerainNet。虽然DerainNet是在合成数据上训练的，但我们发现学习过的网络可以非常有效地转换为真实世界的图像进行测试。此外，我们通过图像增强来增强CNN框架，以改善视觉效果。与现有的单图像去雨方法相比，该方法在网络训练后，具有更好的去雨效果和更快的计算速度。
We introduce a deep network architecture called DerainNet for removing rain streaks from an image. Based on the deep convolutional neural network (CNN), we directly learn the mapping relationship between rainy and clean image detail layers from data. Because we do not possess the ground truth corresponding to real-world rainy images, we synthesize images with rain for training. In contrast to other common strategies that increase depth or breadth of the network, we use image processing domain knowledge to modify the objective function and improve deraining with a modestly sized CNN. Specifically, we train our DerainNet on the detail (high-pass) layer rather than in the image domain. Though DerainNet is trained on synthetic data, we find that the learned network translates very effectively to real-world images for testing. Moreover, we augment the CNN framework with image enhancement to improve the visual results. Compared with the state-of-the-art single image de-raining methods, our method has improved rain removal and much faster computation time after network training.