A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images
This paper proposes a deep convolutional neural network model with encoder-decoder architecture to extract road network from satellite images. We employ ResNet-18 and Atrous Spatial Pyramid Pooling technique to trade off between the extraction precision and running time. A modified cross entropy los...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-05-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/22/5/535 |
| _version_ | 1797568393049538560 |
|---|---|
| author | Bowei Shan Yong Fang |
| author_facet | Bowei Shan Yong Fang |
| author_sort | Bowei Shan |
| collection | DOAJ |
| description | This paper proposes a deep convolutional neural network model with encoder-decoder architecture to extract road network from satellite images. We employ ResNet-18 and Atrous Spatial Pyramid Pooling technique to trade off between the extraction precision and running time. A modified cross entropy loss function is proposed to train our deep model. A PointRend algorithm is used to recover a smooth, clear and sharp road boundary. The augmentated DeepGlobe dataset is used to train our deep model and the asynchronous training method is applied to accelerate the training process. Five salellite images covering Xiaomu village are taken as input to evaluate our model. The proposed E-Road model has fewer number of parameters and shorter training time. The experiments show E-Road outperforms other state-of-the-art deep models with 5.84% to 59.09% improvement, and can give the accurate predictions for the images with complex environment. |
| first_indexed | 2024-03-10T19:55:22Z |
| format | Article |
| id | doaj.art-4980d2c297b94244a91188fb067691ca |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-10T19:55:22Z |
| publishDate | 2020-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-4980d2c297b94244a91188fb067691ca2023-11-19T23:57:29ZengMDPI AGEntropy1099-43002020-05-0122553510.3390/e22050535A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite ImagesBowei Shan0Yong Fang1School of Information Engineering, Chang’an University, Xi’an 710064, ChinaSchool of Information Engineering, Chang’an University, Xi’an 710064, ChinaThis paper proposes a deep convolutional neural network model with encoder-decoder architecture to extract road network from satellite images. We employ ResNet-18 and Atrous Spatial Pyramid Pooling technique to trade off between the extraction precision and running time. A modified cross entropy loss function is proposed to train our deep model. A PointRend algorithm is used to recover a smooth, clear and sharp road boundary. The augmentated DeepGlobe dataset is used to train our deep model and the asynchronous training method is applied to accelerate the training process. Five salellite images covering Xiaomu village are taken as input to evaluate our model. The proposed E-Road model has fewer number of parameters and shorter training time. The experiments show E-Road outperforms other state-of-the-art deep models with 5.84% to 59.09% improvement, and can give the accurate predictions for the images with complex environment.https://www.mdpi.com/1099-4300/22/5/535cross entropyencoder-decoderroad extractiondeep convolutional neural network |
| spellingShingle | Bowei Shan Yong Fang A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images Entropy cross entropy encoder-decoder road extraction deep convolutional neural network |
| title | A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images |
| title_full | A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images |
| title_fullStr | A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images |
| title_full_unstemmed | A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images |
| title_short | A Cross Entropy Based Deep Neural Network Model for Road Extraction from Satellite Images |
| title_sort | cross entropy based deep neural network model for road extraction from satellite images |
| topic | cross entropy encoder-decoder road extraction deep convolutional neural network |
| url | https://www.mdpi.com/1099-4300/22/5/535 |
| work_keys_str_mv | AT boweishan acrossentropybaseddeepneuralnetworkmodelforroadextractionfromsatelliteimages AT yongfang acrossentropybaseddeepneuralnetworkmodelforroadextractionfromsatelliteimages AT boweishan crossentropybaseddeepneuralnetworkmodelforroadextractionfromsatelliteimages AT yongfang crossentropybaseddeepneuralnetworkmodelforroadextractionfromsatelliteimages |