SAR Image Despeckling Based on Block-Matching and Noise-Referenced Deep Learning Method
SAR Image Despeckling Based on Block-Matching and Noise-Referenced Deep Learning Method
复制标题
基于块匹配和噪声参考深度学习方法的SAR图像去斑
DOI：
10.3390/rs14040931
复制
发表时间：
2022-02
期刊：
影响因子：
5
通讯作者：
Zhutao Yang
中科院分区：
文献类型：
--
作者：
Chen Wang;Zhixiang Yin;Xiaoshuang Ma;Zhutao Yang
基于noise 2noise-based的降斑方法能够仅用含噪的合成孔径雷达（SAR）图像训练降斑深度神经网络，在最近的研究中表现出非常好的性能。该方法需要具有较小时间方差的精细注册的多时相数据集，并使用相似性估计来补偿时间方差。然而，构建这样的训练数据集是非常耗时的，并且对于某些从业者来说可能是不可行的。在这篇文章中，我们提出了一种新的单图像散斑方法，该方法结合了基于相似性的块匹配和噪声参考深度学习网络。为该方法设计的去噪网络是一个编码器-解码器卷积神经网络，适用于小图像块。该方法首先在一幅含噪SAR图像或多幅含噪SAR图像中通过基于相似性的块匹配构造大量的含噪图像对作为训练输入。然后，该方法以具有两个参数共享分支的Siamese方式训练网络。与其他最先进的参考滤波器相比，该方法对模拟和真实的SAR数据都具有良好的去斑性能。训练后的网络能够很好地去除同一传感器的不可见图像的斑点噪声，具有良好的泛化能力。所提出的方法的主要优点是其应用的灵活性。它可以用一个有噪声的图像或多个图像进行训练。此外，去斑可以通过专门训练的网络或预先训练的同一传感器来推断。
The noise2noise-based despeckling method, capable of training the despeckling deep neural network with only noisy synthetic aperture radar (SAR) image, has presented very good performance in recent research. This method requires a fine-registered multi-temporal dataset with minor time variance and uses similarity estimation to compensate for the time variance. However, constructing such a training dataset is very time-consuming and may not be viable for a certain practitioner. In this article, we propose a novel single-image-capable speckling method that combines the similarity-based block-matching and noise referenced deep learning network. The denoising network designed for this method is an encoder–decoder convolutional neural network and is accommodated to small image patches. This method firstly constructs a large number of noisy pairs as training input by similarity-based block-matching in either one noisy SAR image or multiple images. Then, the method trains the network in a Siamese manner with two parameter-sharing branches. The proposed method demonstrates favorable despeckling performance with both simulated and real SAR data with respect to other state-of-the-art reference filters. It also presents satisfying generalization capability as the trained network can despeckle well the unseen image of the same sensor. The main advantage of the proposed method is its application flexibility. It could be trained with either one noisy image or multiple images. Furthermore, the despeckling could be inferred by either the ad hoc trained network or a pre-trained one of the same sensor.
登录
查看更多内容
DOI：
10.1109/tgrs.2011.2161586
发表时间：
2012-02-01
影响因子：
8.2
作者：
Parrilli, Sara;Poderico, Mariana;Verdoliva, Luisa
通讯作者：
Verdoliva, Luisa
用于合成孔径雷达图像去斑的深度多尺度循环网络
影响因子：
5
作者：
Yuanyuan Zhou;Jun Shi;Xiaqing Yang;Chen Wang;Durga Kumar;Shunjun Wei;Xiaoling Zhang
通讯作者：
Xiaoling Zhang
DOI：
10.1609/aaai.v33i01.33014031
发表时间：
2019-02
期刊：
ArXiv
影响因子：
--
作者：
Sunghwan Joo;Sungmin Cha;Taesup Moon
通讯作者：
Sunghwan Joo;Sungmin Cha;Taesup Moon
基于像素相关性测量的局部均匀区域分割SAR图像去斑
DOI：
10.1109/tgrs.2011.2107915
发表时间：
2011-07-01
影响因子：
8.2
作者：
Feng, Hongxiao;Hou, Biao;Gong, Maoguo
通讯作者：
Gong, Maoguo
DOI：
10.1109/jstars.2021.3071864
发表时间：
2021
影响因子：
5.5
作者：
Emanuele Dalsasso;L. Denis;F. Tupin
通讯作者：
Emanuele Dalsasso;L. Denis;F. Tupin