GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images
At present, road continuity is a major challenge, and it is difficult to extract the centerline vector of roads, especially when the road view is obstructed by trees or other structures. Most of the existing research has focused on optimizing the available deep-learning networks. However, the segmen...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-05-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/10/2422 |
| _version_ | 1827666721711849472 |
|---|---|
| author | Wei Yuan Wenbo Xu |
| author_facet | Wei Yuan Wenbo Xu |
| author_sort | Wei Yuan |
| collection | DOAJ |
| description | At present, road continuity is a major challenge, and it is difficult to extract the centerline vector of roads, especially when the road view is obstructed by trees or other structures. Most of the existing research has focused on optimizing the available deep-learning networks. However, the segmentation accuracy is also affected by the loss function. Currently, little research has been published on road segmentation loss functions. To resolve this problem, an attention loss function named GapLoss that can be combined with any segmentation network was proposed. Firstly, a deep-learning network was used to obtain a binary prediction mask. Secondly, a vector skeleton was extracted from the prediction mask. Thirdly, for each pixel, eight neighboring pixels with the same value of the pixel were calculated. If the value was 1, then the pixel was identified as the endpoint. Fourth, according to the number of endpoints within a buffered range, each pixel in the prediction image was given a corresponding weight. Finally, the weighted average value of the cross-entropy of all the pixels in the batch was used as the final loss function value. We employed four well-known semantic segmentation networks to conduct comparative experiments on three large datasets. The results showed that, compared to other loss functions, the evaluation metrics after using GapLoss were nearly all improved. From the predicted image, the road prediction by GapLoss was more continuous, especially at intersections and when the road was obscured from view, and the road segmentation accuracy was improved. |
| first_indexed | 2024-03-10T01:56:38Z |
| format | Article |
| id | doaj.art-27d00e54f01a4c688c9798f491953ce4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T01:56:38Z |
| publishDate | 2022-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-27d00e54f01a4c688c9798f491953ce42023-11-23T12:56:00ZengMDPI AGRemote Sensing2072-42922022-05-011410242210.3390/rs14102422GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing ImagesWei Yuan0Wenbo Xu1School of Computer Science, Chengdu University, Chengdu 610106, ChinaSchool of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, ChinaAt present, road continuity is a major challenge, and it is difficult to extract the centerline vector of roads, especially when the road view is obstructed by trees or other structures. Most of the existing research has focused on optimizing the available deep-learning networks. However, the segmentation accuracy is also affected by the loss function. Currently, little research has been published on road segmentation loss functions. To resolve this problem, an attention loss function named GapLoss that can be combined with any segmentation network was proposed. Firstly, a deep-learning network was used to obtain a binary prediction mask. Secondly, a vector skeleton was extracted from the prediction mask. Thirdly, for each pixel, eight neighboring pixels with the same value of the pixel were calculated. If the value was 1, then the pixel was identified as the endpoint. Fourth, according to the number of endpoints within a buffered range, each pixel in the prediction image was given a corresponding weight. Finally, the weighted average value of the cross-entropy of all the pixels in the batch was used as the final loss function value. We employed four well-known semantic segmentation networks to conduct comparative experiments on three large datasets. The results showed that, compared to other loss functions, the evaluation metrics after using GapLoss were nearly all improved. From the predicted image, the road prediction by GapLoss was more continuous, especially at intersections and when the road was obscured from view, and the road segmentation accuracy was improved.https://www.mdpi.com/2072-4292/14/10/2422deep learningloss functionroad extractionremote sensing imagesematic segmentation |
| spellingShingle | Wei Yuan Wenbo Xu GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images Remote Sensing deep learning loss function road extraction remote sensing image sematic segmentation |
| title | GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images |
| title_full | GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images |
| title_fullStr | GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images |
| title_full_unstemmed | GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images |
| title_short | GapLoss: A Loss Function for Semantic Segmentation of Roads in Remote Sensing Images |
| title_sort | gaploss a loss function for semantic segmentation of roads in remote sensing images |
| topic | deep learning loss function road extraction remote sensing image sematic segmentation |
| url | https://www.mdpi.com/2072-4292/14/10/2422 |
| work_keys_str_mv | AT weiyuan gaplossalossfunctionforsemanticsegmentationofroadsinremotesensingimages AT wenboxu gaplossalossfunctionforsemanticsegmentationofroadsinremotesensingimages |