Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5
Lane and object detection are the major concerns of an autonomous vehicle’s ability to move continuously without creating any traffic congestion or collisions. Highly populated rural and urban roads are still facing many challenges to enabling the intelligent transport system with an end-to-end cust...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Zagreb, Faculty of Transport and Traffic Sciences
2025-06-01
|
| Series: | Promet (Zagreb) |
| Subjects: | |
| Online Access: | https://traffic2.fpz.hr/index.php/PROMTT/article/view/669 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850233400094359552 |
|---|---|
| author | Jayamani SIDDAIYAN Kumar PONNUSAMY |
| author_facet | Jayamani SIDDAIYAN Kumar PONNUSAMY |
| author_sort | Jayamani SIDDAIYAN |
| collection | DOAJ |
| description | Lane and object detection are the major concerns of an autonomous vehicle’s ability to move continuously without creating any traffic congestion or collisions. Highly populated rural and urban roads are still facing many challenges to enabling the intelligent transport system with an end-to-end customer connectivity. The proposed work is to identify the drivable space by combining lane line detection by using the LaneNet with sliding window and front road object detection and using the customised YOLOv5. The appropriate pre-processing methods are carried out to reduce the computational complexity and speed up the process. Followed by pre-processing, the reference line is assumed at the far-end distance from the host vehicle to identify the driving space. The lane line borders and objects bounding box coordinate intersecting points on the reference line are picked up to calculate the drivable space. Finally, the proposed system is validated on various public and own datasets. Lane line detection and object detection accuracy of 97% and 98%, respectively are achieved by the LaneNet with sliding windows and custom YOLOv5. |
| format | Article |
| id | doaj-art-4e9f42ebfc5b458daf66d7d4b7ccf2a3 |
| institution | OA Journals |
| issn | 0353-5320 1848-4069 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | University of Zagreb, Faculty of Transport and Traffic Sciences |
| record_format | Article |
| series | Promet (Zagreb) |
| spelling | doaj-art-4e9f42ebfc5b458daf66d7d4b7ccf2a32025-08-20T02:02:57ZengUniversity of Zagreb, Faculty of Transport and Traffic SciencesPromet (Zagreb)0353-53201848-40692025-06-0137373875310.7307/ptt.v37i3.669669Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5Jayamani SIDDAIYAN0Kumar PONNUSAMY1K.S. Rangasamy College of TechnologyK.S. Rangasamy College of TechnologyLane and object detection are the major concerns of an autonomous vehicle’s ability to move continuously without creating any traffic congestion or collisions. Highly populated rural and urban roads are still facing many challenges to enabling the intelligent transport system with an end-to-end customer connectivity. The proposed work is to identify the drivable space by combining lane line detection by using the LaneNet with sliding window and front road object detection and using the customised YOLOv5. The appropriate pre-processing methods are carried out to reduce the computational complexity and speed up the process. Followed by pre-processing, the reference line is assumed at the far-end distance from the host vehicle to identify the driving space. The lane line borders and objects bounding box coordinate intersecting points on the reference line are picked up to calculate the drivable space. Finally, the proposed system is validated on various public and own datasets. Lane line detection and object detection accuracy of 97% and 98%, respectively are achieved by the LaneNet with sliding windows and custom YOLOv5.https://traffic2.fpz.hr/index.php/PROMTT/article/view/669drivable space detectionintelligent vehiclelanenetyolo |
| spellingShingle | Jayamani SIDDAIYAN Kumar PONNUSAMY Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 Promet (Zagreb) drivable space detection intelligent vehicle lanenet yolo |
| title | Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 |
| title_full | Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 |
| title_fullStr | Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 |
| title_full_unstemmed | Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 |
| title_short | Enhancing Autonomous Vehicle Navigation by Detecting Lane and Objects based on LaneNet and CustomYOLOv5 |
| title_sort | enhancing autonomous vehicle navigation by detecting lane and objects based on lanenet and customyolov5 |
| topic | drivable space detection intelligent vehicle lanenet yolo |
| url | https://traffic2.fpz.hr/index.php/PROMTT/article/view/669 |
| work_keys_str_mv | AT jayamanisiddaiyan enhancingautonomousvehiclenavigationbydetectinglaneandobjectsbasedonlanenetandcustomyolov5 AT kumarponnusamy enhancingautonomousvehiclenavigationbydetectinglaneandobjectsbasedonlanenetandcustomyolov5 |