In recent years, convolutional neural networks have achieved considerable success in different computer vision tasks, including image denoising. In this work, we present a residual dense neural network (RDUNet) for image denoising based on the densely connected hierarchical network. The encoding and decoding layers of the RDUNet consist of densely connected convolutional layers to reuse the feature maps and local residual learning to avoid the vanishing gradient problem and speed up the learning process. Moreover, global residual learning is adopted such that, instead of directly predicting the denoised image, the model predicts the residual noise of the corrupted image. The algorithm was trained for the case of additive white Gaussian noise and using a wide range of noise levels. Hence, one advantage of the proposal is that the denoising process does not require prior knowledge about the noise level. In order to evaluate the model, we conducted several experiments with natural image databases available online, achieving competitive results compared with state-of-the-art networks for image denoising. For comparison purpose, we use additive Gaussian noise with levels 10, 30, 50. In the case of grayscale images, we achieved PSNR of 34.39, 29.11, 26.99, and SSIM of 0.9297, 0.8193, 0.7491. For color images we obtained PSNR of 36.68, 31.43, 29.12, and SSIM of 0.9600, 0.8961, 0.8465.

中文翻译：

---

残留密集U-神经网络用于图像去噪

近年来，卷积神经网络已在包括图像去噪在内的各种计算机视觉任务中取得了相当大的成功。在这项工作中，我们提出了一种基于密集连接的层次网络的残差密集神经网络（RDUNet）用于图像去噪。RDUNet的编码和解码层由密集连接的卷积层组成，以重用特征图和局部残差学习，从而避免梯度问题消失并加快学习过程。而且，采用全局残差学习，使得模型预测残差图像的残差噪声，而不是直接预测降噪的图像。该算法针对加性高斯白噪声和广泛的噪声水平进行了训练。因此，该建议的一个优点是，去噪处理不需要关于噪声水平的先验知识。为了评估该模型，我们使用在线可用的自然图像数据库进行了几次实验，与用于图像去噪的最新网络相比，获得了竞争性的结果。为了进行比较，我们使用了10、30、50级的加性高斯噪声。在灰度图像的情况下，我们实现了34.39、29.11、26.99的PSNR和0.9297、0.8193、0.7491的SSIM。对于彩色图像，我们获得的PSNR为36.68、31.43、29.12，SSIM为0.9600、0.8961、0.8465。为了进行比较，我们使用了10、30、50级的加性高斯噪声。在灰度图像的情况下，我们实现了34.39、29.11、26.99的PSNR和0.9297、0.8193、0.7491的SSIM。对于彩色图像，我们获得的PSNR为36.68、31.43、29.12，SSIM为0.9600、0.8961、0.8465。为了进行比较，我们使用了10、30、50级的加性高斯噪声。在灰度图像的情况下，我们实现了34.39、29.11、26.99的PSNR和0.9297、0.8193、0.7491的SSIM。对于彩色图像，我们获得的PSNR为36.68、31.43、29.12，SSIM为0.9600、0.8961、0.8465。