Remote sensing image denoising using residual encoder-decoder networks with edge enhancement

Congzhong WU; Xi CHEN; Shu ZHAN

doi:10.11834/jrs.20208169

Chinese Satellites | Views : 0 下载量: 160 CSCD: 5 更多指标

PDF
Export
Share
Collection
Album

Remote sensing image denoising using residual encoder-decoder networks with edge enhancement
Vol. 24, Issue 1, Pages: 27-36(2020)
Published： 07 January 2020 ，
DOI： 10.11834/jrs.20208169

扫描看全文

吴从中,陈曦,詹曙.2020.结合残差编解码网络和边缘增强的遥感图像去噪.遥感学报,24(1): 27-36

WU Congzhong,CHEN Xi,ZHAN Shu. 2020. Remote sensing image denoising using residual encoder-decoder networks with edge enhancement. Journal of Remote Sensing(Chinese). 24(1): 27-36
吴从中,陈曦,詹曙.2020.结合残差编解码网络和边缘增强的遥感图像去噪.遥感学报,24(1): 27-36 DOI： 10.11834/jrs.20208169.

WU Congzhong,CHEN Xi,ZHAN Shu. 2020. Remote sensing image denoising using residual encoder-decoder networks with edge enhancement. Journal of Remote Sensing(Chinese). 24(1): 27-36 DOI： 10.11834/jrs.20208169.

摘要

高分辨率遥感图像去噪对于提高后续图像分析、识别等问题的准确性具有重要意义。目前的去噪算法普遍存在去噪结果边缘信息模糊、易产生视觉伪影导致遥感信息丢失的缺点，针对以上问题本文提出了一个基于边缘增强的残差编解码去噪网络用于高分辨率遥感影像去噪。该方法首先将噪声图片通过低通滤波器分解成高频层和低频层，然后将含噪声信息的高频层输入到带残差模块的编解码网络中，通过采样运算在多尺度空间上学习残差映射生成残差图像，最后使用跳跃连接得到完整的去噪结果。其损失函数由逐像素和感知损失两部分组成，逐像素损失使用传统的均方根误差学习像素级信息，感知损失学习语义特征上的差异可以保留更多边缘信息，最终得到更清晰的结果，其中感知损失是由级联在后的语义分割网络提取的特征图定义的。本文对不同测试数据做去噪实验并与几个经典方法对比证明本文模型的去噪结果优于其他方法，不仅提高图像的峰值信噪比，得到最高的平均梯度值，还在视觉上取得了最清晰的结果。实验结果表明，本文提出的基于边缘增强的深层编解码卷积网络在去噪的同时可以改善边缘细节被模糊的问题，保留更多遥感地物信息，提高图像视觉效果。

Abstract

Remote sensing images are often affected by noise in the process of digitization and transmission processes. Denoising is an indispensable way of improving image quality. Despite showing an excellent noise removal performance

the existing denoising algorithms however typically suffer from a common drawback. Specifically

in the learning process

some edge information is lost

thereby over-smoothing the denoising result. Given the importance of details—including sharp edges and texture information—in remote sensing images

we propose a residual encoder–decoder denoising network with joint loss (REDJ) for GF-2 satellite data.

Inspired by U-net

we use a deep convolutional framework is used to learn the end-to-end mapping from noisy images to the original ones. The encoder acts as a feature extractor that captures semantic information of image contents while eliminating noise

whereas the decoder recovers the image details. The high-resolution features from the encoder are combined with the up-sampled output by skip connection. We also introduce high-frequency decomposition and residual mapping to simplify the training process by reducing the solution space. As for the loss function

we modify the traditional denoising per-pixel loss. Given a well-trained convolutional neural network for defining perceptual loss

we instead to learn the perceptual differences of the extracted features instead of merely matching the low-level pixel information. Unlike the loss of detail resulting from normal per-pixel MSE loss

we recommend a new joint loss that combines the advantages of both per-pixel reconstruction and feature reconstruction

preserves additional edge and texture information

and generates clear denoised results. We employ the GF-2 satellite images in the experiments. To obtain enough training and testing data

we divide the entire high-resolution image is divided into 1200 pictures of size 512 and then allocate 70% of these images for training and the other 30% for testing. We generate the noisy images by adding Gaussian noise.

To verify the effectiveness of our proposed network

we compare our quantitative and qualitative results with those of other state-of-the-art methods

including wavelet threshold

total variation

and K-SVD. Our proposed method REDJ can obtain the best index values both of PSNR and average gradient. In the qualitative visual sense

REDJ obtains clear denoising results because of the joint of perceptual loss. Compared with other methods that produce blurred regions generated by other methods

REDJ preserves more edge information and texture details. We also compare the run times of different methods for denoising images and find that REDJ has a relatively high CPU speed and achieves an excellent computational efficiency on GPU time.

This paper successfully applies deep learning theory for denoising remote sensing images. We use the proposed network is used to remove noise from high-resolution GF-2 remote sensing images and to preserve the edge contours and fine details

which is conducive to facilitate later detection

classification

and other remote sensing applications. In our future work

we will explore to handle other types of noises

especially the complex real-world noises

and consider a single comprehensive network for more image restoration tasks.

关键词

遥感图像去噪卷积神经网络边缘增强感知损失高分二号

Keywords

remote sensing image denoisingconvolutional neural networkedge enhancementperceptual lossGF-2 remote sensing image

references

Badrinarayanan V, Kendall A and Cipolla R. 2017. SegNet: a deep convolutional encoder-decoder architecture for image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12): 2481-2495 [DOI: 10.1109/TPAMI.2016.2644615http://dx.doi.org/10.1109/TPAMI.2016.2644615]

Chang Y, Yan L X, Fang H Z and Liu H. 2014. Simultaneous destriping and denoising for remote sensing images with unidirectional total variation and sparse representation. IEEE Geoscience and Remote Sensing Letters, 11(6): 1051-1055 [DOI: 10.1109/LGRS.2013.2285124http://dx.doi.org/10.1109/LGRS.2013.2285124]

Ghassemian H. 2016. A review of remote sensing image fusion methods. Information Fusion, 32: 75-89 [DOI: 10.1016/j.inffus.2016.03.003http://dx.doi.org/10.1016/j.inffus.2016.03.003]

He K M, Sun J and Tang X O. 2013. Guided image filtering. IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(6): 1397-1409 [DOI: 10.1109/TPAMI.2012.213http://dx.doi.org/10.1109/TPAMI.2012.213]

He K M, Zhang X Y, Ren S Q and Sun J. 2016. Deep residual learning for image recognition//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE: 770-778 [DOI: 10.1109/CVPR.2016.90http://dx.doi.org/10.1109/CVPR.2016.90]

He W, Zhang H Y, Zhang L P and Shen H F. 2016. Total-variation-regularized low-rank matrix factorization for hyperspectral image restoration. IEEE Transactions on Geoscience and Remote Sensing, 54(1): 178-188 [DOI: 10.1109/TGRS.2015.2452812http://dx.doi.org/10.1109/TGRS.2015.2452812]

Jain V and Seung H S. 2008. Natural image denoising with convolutional networks//Proceedings of the 21st International Conference on Neural Information Processing Systems. Vancouver, British Columbia, Canada: Curran Associates Inc.: 769-776

Johnson J, Alahi A and Li F F. 2016. Perceptual losses for real-time style transfer and super-resolution//European Conference on Computer Vision. Cham: Springer: 694-711 [DOI: 10.1007/978-3-319-46475-6_43http://dx.doi.org/10.1007/978-3-319-46475-6_43]

Kim J, Lee J K and Lee K M. 2016. Accurate image super-resolution using very deep convolutional networks// 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE: 1646-1654 [DOI: 10.1109/CVPR.2016.182http://dx.doi.org/10.1109/CVPR.2016.182]

Ledig C, Theis L, Huszár F, Caballero J, Cunningham A, Acosta A, Aitken A, Tejani A, Totz J, Wang Z H and Shi W Z. 2017. Photo-realistic single image super-resolution using a generative adversarial network//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, USA: IEEE: 105-114 [DOI: 10.1109/CVPR.2017.19http://dx.doi.org/10.1109/CVPR.2017.19]

Li J, Yuan Q Q, Shen H F and Zhang L P. 2015. Hyperspectral image recovery employing a multidimensional nonlocal total variation model. Signal Processing, 111: 230-248 [DOI: 10.1016/j.sigpro.2014.12.023http://dx.doi.org/10.1016/j.sigpro.2014.12.023]

Liu X X, Lu X L, Shen H F, Yuan Q Q, Jiao Y L and Zhang L P. 2016. Stripe noise separation and removal in remote sensing images by consideration of the global sparsity and local variational properties. IEEE Transactions on Geoscience and Remote Sensing, 54(5): 3049-3060 [DOI: 10.1109/TGRS.2015.2510418http://dx.doi.org/10.1109/TGRS.2015.2510418]

Mao X J, Shen C H and Yang Y B. 2016. Image restoration using convolutional auto-encoders with symmetric skip connections. arXiv:1606.08921

Ronneberger O, Fischer P and Brox T. 2015. U-Net: convolutional networks for biomedical image segmentation//Navab N, Hornegger J, Wells W and Frangi A, eds. International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer: 234-241 [DOI: 10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28]

Vincent P, Larochelle H, Bengio Y and Manzagol P A. 2008. Extracting and composing robust features with denoising autoencoders//Proceedings of the 25th International Conference on Machine Learning. Helsinki, Finland: ACM: 1096-1103 [DOI: 10.1145/1390156.1390294http://dx.doi.org/10.1145/1390156.1390294]

Xia Q, Xing S, Ma D Y, Mo D L, Li P C and Ge Z X. 2016. An improved K-SVD-based denoising method for remote sensing satellite images. Journal of Remote Sensing, 20(3): 441-449

夏琴, 邢帅, 马东洋, 莫德林, 李鹏程, 葛忠孝. 2016. 遥感卫星影像K-SVD稀疏表示去噪. 遥感学报, 20(3): 441-449 [DOI: 10.11834/jrs.20165149http://dx.doi.org/10.11834/jrs.20165149]

Xie J Y, Xu L L and Chen E H. 2012. Image denoising and inpainting with deep neural networks//International Conference on Neural Information Processing Systems. Lake Tahoe, Nevada: Curran Associates Inc.: 341-349

Xu D, Sun L and Luo J S. 2015. Denoising of hyperspectral remote sensing imagery using NAPCA and complex wavelet transform. Infrared and Laser Engineering, 44(1): 327-334

徐冬, 孙蕾, 罗建书. 2015. 结合NAPCA和复小波变换的高光谱遥感图像去噪. 红外与激光工程, 44(1): 327-334 [DOI: 10.3969/j.issn.1007-2276.2015.01.056http://dx.doi.org/10.3969/j.issn.1007-2276.2015.01.056]

Zhang C M, Sun Q S, Wang C, Feng L and Gu Y H. 2016. Remote sensing image de-noising via sparse representation based on non-local self-similarity. Remote Sensing Technology and Application, 31(4): 739-747

张从梅, 孙权森, 王超, 封磊, 顾一禾. 2016. 基于非局部自相似性的遥感图像稀疏去噪算法. 遥感技术与应用, 31(4): 739-747[DOI: 10.11873/j.issn.1004-0323.2016.4.0739http://dx.doi.org/10.11873/j.issn.1004-0323.2016.4.0739]

Alert me when the article has been cited

提交

MtSCCD： Land-use scene classification and change-detection dataset for deep learning

Hyperspectral unmixing network considering spectral information and superpixel segmentation

Cultivated land extraction from high-resolution remote sensing images based on BECU-Net model with edge enhancement

Spectral-spatial attention bilateral network for hyperspectral image classification

Automatic detection method of a moving ship based on an HY-1/CZI satellite image