Road Extraction by Deep Residual U-Net
By Zhengxin Zhang, Qingjie Liu, Yunhong Wang. IEEE Geoscience and Remote Sensing Letters, 2017
2,740 citations recorded by the source corpus.
Road extraction from aerial images has been a hot research topic in the field of remote sensing image analysis. In this letter, a semantic segmentation neural network, which combines the strengths of residual learning and U-Net, is proposed for road area extraction. The network is built with residual units and has similar architecture to that of U-Net. The benefits of this model are twofold: first, residual units ease training of deep networks. Second, the rich skip connections within the network could facilitate information propagation, allowing us to design networks with fewer parameters, however, better performance. We test our network on a public road data set and compare it with U-Net and other two state-of-the-art deep-learning-based road extraction methods. The proposed approach outperforms all the comparing methods, which demonstrates its superiority over recently developed state of the arts.
Subfields: Environmental Engineering, Media Technology, Ocean Engineering
View Paper: https://www.semanticscholar.org/paper/f06ff5f719eb9cd939dde8fc9b199b17adcbc75f
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Environmental Engineering
- Michael Alexander Riegler
- Pål Halvorsen
- Antonio J. Plaza
- Thomas de Lange
- Chuang Zhang
- Jeya Maria Jose Valanarasu
- Matteo Matteucci
- Ming Wu
In Media Technology
- Radu Timofte
- Kai Zhang
- Luc Van Gool
- Wangmeng Zuo
- Lei Zhang
- Chao Dong
- Chen Change Loy
- Ming-Hsuan Yang
In Ocean Engineering
- Vladimir I. Iglovikov
- Michael Alexander Riegler
- Pål Halvorsen
- Antonio J. Plaza
- Thomas de Lange
- Chuang Zhang
- Jeya Maria Jose Valanarasu
- Matteo Matteucci
Citations in the corpus
- CaraNet: context axial reverse attention network for segmentation of small medical objects (2021, methodology): “We compare our CaraNet with six state-of-the-art medical image segmentation methods: UNet [1], U-Net++ [2], ResUNetmod [15], ResUNet++ [3], SFA [16] and PraNet [4].”
- Medical Image Segmentation Review: The Success of U-Net (2022, methodology): “Residual U-Net [173], even with the Residual enhancements, is no exception to this rule.”
- Fully Convolutional Network for Automatic Road Extraction from Satellite Imagery (2018, background): “This type of network architecture proven themselves in binary image segmentation competitions for satellite image analyses [6, 7, 8, 15, 16] and other applications [13, 12].”
- A review of deep learning methods for semantic segmentation of remote sensing imagery (2021, background): “Xu, Du, Zhang, and Zhang (2018) proposed a Random Patches Network that takes image patches as the convolution kernels, which requires no training.”
- TransBTS: Multimodal Brain Tumor Segmentation Using Transformer (2021, neutral): “as U-Net++ [21] and Res-UNet [20] further improve the performance for image segmentation.”
- Plug-and-Play Image Restoration With Deep Denoiser Prior (2020, result): “Actually, the similar idea of combining U-Net and ResNet can also be found in other works such as [61], [62].”
- DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation (2020, neutral): “[15] proposed Deep Residual U-Net, which uses residual connections better output segmentation map.”
- DeepCMB: Lensing Reconstruction of the Cosmic Microwave Background with Deep Neural Networks (2018, methodology): “We employ an instance from a family of network architectures, ResUNets (Kayalıbay et al., 2017; Zhang et al., 2017), which learns a transformation between images.”
- ResUNet++: An Advanced Architecture for Medical Image Segmentation (2019, methodology): “There are various modification and extension based on U-Net architecture [4], [6], [17], [19]–[22] to achieve better segmentation results on both natural images and biomedical images.”
- Kvasir-SEG: A Segmented Polyp Dataset (2019, methodology): “ conducted initial investigations using two dierent methods. The rst method is based on the ecient FCM [6] unsupervised clustering algorithm. The second method is based on the deep-learning ResUNet [28] architecture, utilizing the advantage of the residual block. When using basic CNN architectures t”
- Deep Learning for Land Use and Land Cover Classification Based on Hyperspectral and Multispectral Earth Observation Data: A Review (2020, background): “In the context of remote sensing, U-Net has shown very promising results in extracting buildings [127,128], roads [129,130], clouds [131,132], and to classify other land covers [133–135] using high resolution MSI data.”
- INet: Convolutional Networks for Biomedical Image Segmentation (2021, methodology): “Compared to the commonly used baseline model, UNet, the proposed INet has additional residual shortcuts, and therefore the learning task of the optimizer becomes easier [39] (We denote INet without customized residual short-cuts and the Conv-index as Plain-INet).”
- Very high resolution canopy height maps from RGB imagery using self-supervised vision transformer and convolutional decoder trained on aerial lidar (2023, methodology): “We utilized a ResUNet-18 architecture for our baseline (Zhang et al., 2017), which is an encoder-decoder architecture predicting a N × N canopy height map from a 3 × N × N RGB image, with N = 256.”
- DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation (2022, methodology): “The VGGNet [39] and typical residual structures [23] have been applied in many previous semantic segmentation networks, such as DoubleUnet [22] and ResUnet [40].”
- UNeXt: MLP-based Rapid Medical Image Segmentation Network (2022, methodology): “In particular, we compare with convolutional baselines like UNet [17], UNet++ [29] and ResUNet [28].”
- COVID-CT-Dataset: A CT Scan Dataset about COVID-19 (2020, methodology): “Using the labeled lung masks in COVID-CT and COVID-Seg, we develop a lung segmentation model based on residual U-Net (Zhang et al., 2018).”
- Recurrent Residual Convolutional Neural Network based on U-Net (R2U-Net) for Medical Image Segmentation (2018, methodology): “However, U-Net is not only limited to the applications in the domain of medical imaging, nowadays this model is massively applied for computer vision tasks as well [30, 31].”