Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields
Road extraction from remote sensing images is of great significance to urban planning, navigation, disaster assessment, and other applications. Although deep neural networks have shown a strong ability in road extraction, it remains a challenging task due to complex circumstances and factors such as...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-01-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/3/465 |
| _version_ | 1797406206190419968 |
|---|---|
| author | Shuyang Wang Xiaodong Mu Dongfang Yang Hao He Peng Zhao |
| author_facet | Shuyang Wang Xiaodong Mu Dongfang Yang Hao He Peng Zhao |
| author_sort | Shuyang Wang |
| collection | DOAJ |
| description | Road extraction from remote sensing images is of great significance to urban planning, navigation, disaster assessment, and other applications. Although deep neural networks have shown a strong ability in road extraction, it remains a challenging task due to complex circumstances and factors such as occlusion. To improve the accuracy and connectivity of road extraction, we propose an inner convolution integrated encoder-decoder network with the post-processing of directional conditional random fields. Firstly, we design an inner convolutional network which can propagate information slice-by-slice within feature maps, thus enhancing the learning of road topology and linear features. Additionally, we present the directional conditional random fields to improve the quality of the extracted road by adding the direction of roads to the energy function of the conditional random fields. The experimental results on the Massachusetts road dataset show that the proposed approach achieves high-quality segmentation results, with the F1-score of 84.6%, which outperforms other comparable “state-of-the-art” approaches. The visualization results prove that the proposed approach is able to effectively extract roads from remote sensing images and can solve the road connectivity problem produced by occlusions to some extent. |
| first_indexed | 2024-03-09T03:23:01Z |
| format | Article |
| id | doaj.art-f43544ea673a41c198f799f18de328f7 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T03:23:01Z |
| publishDate | 2021-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-f43544ea673a41c198f799f18de328f72023-12-03T15:07:28ZengMDPI AGRemote Sensing2072-42922021-01-0113346510.3390/rs13030465Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random FieldsShuyang Wang0Xiaodong Mu1Dongfang Yang2Hao He3Peng Zhao4Department of Information Engineering, Xi’an Hi-Tech Research Institute, Xi’an 710025, ChinaDepartment of Information Engineering, Xi’an Hi-Tech Research Institute, Xi’an 710025, ChinaDepartment of Control Engineering, Xi’an Hi-Tech Research Institute, Xi’an 710025, ChinaDepartment of Measurement and Control, Qingzhou Hi-Tech Research Institute, Qingzhou 262500, ChinaDepartment of Information Engineering, Xi’an Hi-Tech Research Institute, Xi’an 710025, ChinaRoad extraction from remote sensing images is of great significance to urban planning, navigation, disaster assessment, and other applications. Although deep neural networks have shown a strong ability in road extraction, it remains a challenging task due to complex circumstances and factors such as occlusion. To improve the accuracy and connectivity of road extraction, we propose an inner convolution integrated encoder-decoder network with the post-processing of directional conditional random fields. Firstly, we design an inner convolutional network which can propagate information slice-by-slice within feature maps, thus enhancing the learning of road topology and linear features. Additionally, we present the directional conditional random fields to improve the quality of the extracted road by adding the direction of roads to the energy function of the conditional random fields. The experimental results on the Massachusetts road dataset show that the proposed approach achieves high-quality segmentation results, with the F1-score of 84.6%, which outperforms other comparable “state-of-the-art” approaches. The visualization results prove that the proposed approach is able to effectively extract roads from remote sensing images and can solve the road connectivity problem produced by occlusions to some extent.https://www.mdpi.com/2072-4292/13/3/465remote sensingsemantic segmentationencoder-decoder networkinner convolutiondirectional conditional random fields |
| spellingShingle | Shuyang Wang Xiaodong Mu Dongfang Yang Hao He Peng Zhao Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields Remote Sensing remote sensing semantic segmentation encoder-decoder network inner convolution directional conditional random fields |
| title | Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields |
| title_full | Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields |
| title_fullStr | Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields |
| title_full_unstemmed | Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields |
| title_short | Road Extraction from Remote Sensing Images Using the Inner Convolution Integrated Encoder-Decoder Network and Directional Conditional Random Fields |
| title_sort | road extraction from remote sensing images using the inner convolution integrated encoder decoder network and directional conditional random fields |
| topic | remote sensing semantic segmentation encoder-decoder network inner convolution directional conditional random fields |
| url | https://www.mdpi.com/2072-4292/13/3/465 |
| work_keys_str_mv | AT shuyangwang roadextractionfromremotesensingimagesusingtheinnerconvolutionintegratedencoderdecodernetworkanddirectionalconditionalrandomfields AT xiaodongmu roadextractionfromremotesensingimagesusingtheinnerconvolutionintegratedencoderdecodernetworkanddirectionalconditionalrandomfields AT dongfangyang roadextractionfromremotesensingimagesusingtheinnerconvolutionintegratedencoderdecodernetworkanddirectionalconditionalrandomfields AT haohe roadextractionfromremotesensingimagesusingtheinnerconvolutionintegratedencoderdecodernetworkanddirectionalconditionalrandomfields AT pengzhao roadextractionfromremotesensingimagesusingtheinnerconvolutionintegratedencoderdecodernetworkanddirectionalconditionalrandomfields |