Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset
Road markings, including road lanes and symbolic road markings, can convey abundant guidance information to autonomous driving cars. However, recent works have paid less attention to the recognition of symbolic road markings compared with road lanes. In this study, a road-marking-segmentation datase...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/18/4508 |
| _version_ | 1827657213097803776 |
|---|---|
| author | Junjie Wu Wen Liu Yoshihisa Maruyama |
| author_facet | Junjie Wu Wen Liu Yoshihisa Maruyama |
| author_sort | Junjie Wu |
| collection | DOAJ |
| description | Road markings, including road lanes and symbolic road markings, can convey abundant guidance information to autonomous driving cars. However, recent works have paid less attention to the recognition of symbolic road markings compared with road lanes. In this study, a road-marking-segmentation dataset named the RMD (Road Marking Dataset) is introduced to compensate for the lack of datasets and the limitations of the existing datasets. Furthermore, we propose a novel multiscale attention-based dilated convolutional neural network (MSA-DCNN) to tackle the proposed RMD. The proposed method employs multiscale attention to merge the weighting outputs of adjacent multiscale inputs, and dilated convolution to capture spatial-context information. The performance analysis shows that the proposed MSA-DCNN yields the best results by combining multiscale attention and dilated convolution. Additionally, the proposed method gains the mIoU of 74.88%, which is a significant improvement over the existing techniques. |
| first_indexed | 2024-03-09T22:39:09Z |
| format | Article |
| id | doaj.art-e213495ee164474ea33aab9f889a7b46 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T22:39:09Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-e213495ee164474ea33aab9f889a7b462023-11-23T18:43:43ZengMDPI AGRemote Sensing2072-42922022-09-011418450810.3390/rs14184508Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking DatasetJunjie Wu0Wen Liu1Yoshihisa Maruyama2Graduate School of Engineering, Chiba University, Inage-ku, Chiba 263-8522, JapanGraduate School of Engineering, Chiba University, Inage-ku, Chiba 263-8522, JapanGraduate School of Engineering, Chiba University, Inage-ku, Chiba 263-8522, JapanRoad markings, including road lanes and symbolic road markings, can convey abundant guidance information to autonomous driving cars. However, recent works have paid less attention to the recognition of symbolic road markings compared with road lanes. In this study, a road-marking-segmentation dataset named the RMD (Road Marking Dataset) is introduced to compensate for the lack of datasets and the limitations of the existing datasets. Furthermore, we propose a novel multiscale attention-based dilated convolutional neural network (MSA-DCNN) to tackle the proposed RMD. The proposed method employs multiscale attention to merge the weighting outputs of adjacent multiscale inputs, and dilated convolution to capture spatial-context information. The performance analysis shows that the proposed MSA-DCNN yields the best results by combining multiscale attention and dilated convolution. Additionally, the proposed method gains the mIoU of 74.88%, which is a significant improvement over the existing techniques.https://www.mdpi.com/2072-4292/14/18/4508road-marking segmentationmultiscale attentiondilated convolutiondeep learning |
| spellingShingle | Junjie Wu Wen Liu Yoshihisa Maruyama Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset Remote Sensing road-marking segmentation multiscale attention dilated convolution deep learning |
| title | Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset |
| title_full | Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset |
| title_fullStr | Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset |
| title_full_unstemmed | Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset |
| title_short | Automated Road-Marking Segmentation via a Multiscale Attention-Based Dilated Convolutional Neural Network Using the Road Marking Dataset |
| title_sort | automated road marking segmentation via a multiscale attention based dilated convolutional neural network using the road marking dataset |
| topic | road-marking segmentation multiscale attention dilated convolution deep learning |
| url | https://www.mdpi.com/2072-4292/14/18/4508 |
| work_keys_str_mv | AT junjiewu automatedroadmarkingsegmentationviaamultiscaleattentionbaseddilatedconvolutionalneuralnetworkusingtheroadmarkingdataset AT wenliu automatedroadmarkingsegmentationviaamultiscaleattentionbaseddilatedconvolutionalneuralnetworkusingtheroadmarkingdataset AT yoshihisamaruyama automatedroadmarkingsegmentationviaamultiscaleattentionbaseddilatedconvolutionalneuralnetworkusingtheroadmarkingdataset |