Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network
Most lane line detection algorithms still have room for improvement in detection accuracy, speed, and robustness. Meanwhile, these algorithms only test the performance indicators through the test set of the open-source dataset rather than deploying them on actual vehicles and evaluating the performa...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2022-12-01
|
| Series: | Applied Artificial Intelligence |
| Online Access: | http://dx.doi.org/10.1080/08839514.2022.2085321 |
| _version_ | 1827774589535518720 |
|---|---|
| author | Sun Yang Li Yunpeng Liu Yu |
| author_facet | Sun Yang Li Yunpeng Liu Yu |
| author_sort | Sun Yang |
| collection | DOAJ |
| description | Most lane line detection algorithms still have room for improvement in detection accuracy, speed, and robustness. Meanwhile, these algorithms only test the performance indicators through the test set of the open-source dataset rather than deploying them on actual vehicles and evaluating the performance indicators through road scenarios. Therefore, this paper proposes a lane detection algorithm based on instance segmentation. Firstly, a dual-branch neural network model for lane line image segmentation was designed based on BiSeNet V2. Then the discrete lane line feature points are operated through the clustering model. The corresponding feature points are selected for fitting by combining straight lines and curves to obtain the appropriate fitting parameter equation for the specific visual field area. Finally, the model is trained and verified based on the TuSimple dataset. The algorithm has a noticeable performance improvement under the two evaluation indicators of mIoU and FPS. Meanwhile, the model is integrated into the ROS task platform for intelligent vehicles. The results show that the algorithm’s accuracy and detection speed are increased to about 3.9 and 2.9 times, respectively, that of the improved probabilistic Hough transform algorithm under the two evaluation indicators of lateral distance and the detection time of each image frame. |
| first_indexed | 2024-03-11T13:40:29Z |
| format | Article |
| id | doaj.art-8a8e1dd4db89445e8b36c3c83c4a5ddd |
| institution | Directory Open Access Journal |
| issn | 0883-9514 1087-6545 |
| language | English |
| last_indexed | 2024-03-11T13:40:29Z |
| publishDate | 2022-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Applied Artificial Intelligence |
| spelling | doaj.art-8a8e1dd4db89445e8b36c3c83c4a5ddd2023-11-02T13:36:38ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452022-12-0136110.1080/08839514.2022.20853212085321Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone NetworkSun Yang0Li Yunpeng1Liu Yu2Hebei University of Engineering, and the Handan Key Laboratory of Intelligent VehiclesHebei University of EngineeringHebei University of EngineeringMost lane line detection algorithms still have room for improvement in detection accuracy, speed, and robustness. Meanwhile, these algorithms only test the performance indicators through the test set of the open-source dataset rather than deploying them on actual vehicles and evaluating the performance indicators through road scenarios. Therefore, this paper proposes a lane detection algorithm based on instance segmentation. Firstly, a dual-branch neural network model for lane line image segmentation was designed based on BiSeNet V2. Then the discrete lane line feature points are operated through the clustering model. The corresponding feature points are selected for fitting by combining straight lines and curves to obtain the appropriate fitting parameter equation for the specific visual field area. Finally, the model is trained and verified based on the TuSimple dataset. The algorithm has a noticeable performance improvement under the two evaluation indicators of mIoU and FPS. Meanwhile, the model is integrated into the ROS task platform for intelligent vehicles. The results show that the algorithm’s accuracy and detection speed are increased to about 3.9 and 2.9 times, respectively, that of the improved probabilistic Hough transform algorithm under the two evaluation indicators of lateral distance and the detection time of each image frame.http://dx.doi.org/10.1080/08839514.2022.2085321 |
| spellingShingle | Sun Yang Li Yunpeng Liu Yu Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network Applied Artificial Intelligence |
| title | Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network |
| title_full | Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network |
| title_fullStr | Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network |
| title_full_unstemmed | Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network |
| title_short | Lane Detection Based on Instance Segmentation of BiSeNet V2 Backbone Network |
| title_sort | lane detection based on instance segmentation of bisenet v2 backbone network |
| url | http://dx.doi.org/10.1080/08839514.2022.2085321 |
| work_keys_str_mv | AT sunyang lanedetectionbasedoninstancesegmentationofbisenetv2backbonenetwork AT liyunpeng lanedetectionbasedoninstancesegmentationofbisenetv2backbonenetwork AT liuyu lanedetectionbasedoninstancesegmentationofbisenetv2backbonenetwork |