A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs
It is a challenging task for self-driving vehicles in Real-World traffic scenarios to find a trade-off between the real-time performance and the high accuracy of the detection, recognition, and tracking in videos. This issue is addressed in this paper with an improved YOLOv3 (You Only Look Once) and...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-03-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/11/7/3061 |
| _version_ | 1797539541728362496 |
|---|---|
| author | Shaojian Song Yuanchao Li Qingbao Huang Gang Li |
| author_facet | Shaojian Song Yuanchao Li Qingbao Huang Gang Li |
| author_sort | Shaojian Song |
| collection | DOAJ |
| description | It is a challenging task for self-driving vehicles in Real-World traffic scenarios to find a trade-off between the real-time performance and the high accuracy of the detection, recognition, and tracking in videos. This issue is addressed in this paper with an improved YOLOv3 (You Only Look Once) and a multi-object tracking algorithm (Deep-Sort). First, data augmentation is employed for small sample traffic signs to address the problem of an extremely unbalanced distribution of different samples in the dataset. Second, a new architecture of YOLOv3 is proposed to make it more suitable for detecting small targets. The detailed method is (1) removing the output feature map corresponding to the 32-times subsampling of the input image in the original YOLOv3 structure to reduce its computational costs and improve its real-time performances; (2) adding an output feature map of 4-times subsampling to improve its detection capability for the small traffic signs; (3) Deep-Sort is integrated into the detection method to improve the precision and robustness of multi-object detection, and the tracking ability in videos. Finally, our method demonstrated better detection capabilities, with respect to state-of-the-art approaches, which precision, recall and mAP is 91%, 90%, and 84.76% respectively. |
| first_indexed | 2024-03-10T12:47:17Z |
| format | Article |
| id | doaj.art-4b85e51f80924714ab3bf0c1f36e10c1 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T12:47:17Z |
| publishDate | 2021-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-4b85e51f80924714ab3bf0c1f36e10c12023-11-21T13:20:57ZengMDPI AGApplied Sciences2076-34172021-03-01117306110.3390/app11073061A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic SignsShaojian Song0Yuanchao Li1Qingbao Huang2Gang Li3School of Electrical Engineering, Guangxi University, Nanning 530004, ChinaSchool of Electrical Engineering, Guangxi University, Nanning 530004, ChinaSchool of Electrical Engineering, Guangxi University, Nanning 530004, ChinaSchool of Electrical Engineering, Guangxi University, Nanning 530004, ChinaIt is a challenging task for self-driving vehicles in Real-World traffic scenarios to find a trade-off between the real-time performance and the high accuracy of the detection, recognition, and tracking in videos. This issue is addressed in this paper with an improved YOLOv3 (You Only Look Once) and a multi-object tracking algorithm (Deep-Sort). First, data augmentation is employed for small sample traffic signs to address the problem of an extremely unbalanced distribution of different samples in the dataset. Second, a new architecture of YOLOv3 is proposed to make it more suitable for detecting small targets. The detailed method is (1) removing the output feature map corresponding to the 32-times subsampling of the input image in the original YOLOv3 structure to reduce its computational costs and improve its real-time performances; (2) adding an output feature map of 4-times subsampling to improve its detection capability for the small traffic signs; (3) Deep-Sort is integrated into the detection method to improve the precision and robustness of multi-object detection, and the tracking ability in videos. Finally, our method demonstrated better detection capabilities, with respect to state-of-the-art approaches, which precision, recall and mAP is 91%, 90%, and 84.76% respectively.https://www.mdpi.com/2076-3417/11/7/3061object detectionmulti-object trackingimproved YOLOv3deep learningself-driving vehicles |
| spellingShingle | Shaojian Song Yuanchao Li Qingbao Huang Gang Li A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs Applied Sciences object detection multi-object tracking improved YOLOv3 deep learning self-driving vehicles |
| title | A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs |
| title_full | A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs |
| title_fullStr | A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs |
| title_full_unstemmed | A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs |
| title_short | A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs |
| title_sort | new real time detection and tracking method in videos for small target traffic signs |
| topic | object detection multi-object tracking improved YOLOv3 deep learning self-driving vehicles |
| url | https://www.mdpi.com/2076-3417/11/7/3061 |
| work_keys_str_mv | AT shaojiansong anewrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT yuanchaoli anewrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT qingbaohuang anewrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT gangli anewrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT shaojiansong newrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT yuanchaoli newrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT qingbaohuang newrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns AT gangli newrealtimedetectionandtrackingmethodinvideosforsmalltargettrafficsigns |