To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Road Traffic Vehicle Detection and Tracking using Deep Learning with Custom-Collected and Public Datasets |
|
Author |
Anwaar Alshareef, Aiiad Albeshri, Iyad Katib, and Rashid Mehmood |
| Citation |
Vol. 20 No. 11 pp. 9-21 |
|
Abstract |
Deep learning is revolutionizing smart cities and societies, solving many longstanding problems. Transportation is continuing to cause unbelievable damages including 1.25 million deaths and trillions of dollars annually. This paper presents a study on the use of YOLOv4 for vehicle detection and DeepSORT for tracking the detected vehicles on roads. None of the earlier works have applied these models to road traffic in the Kingdom of Saudi Arabia (KSA). We have used three different variations of the deep learning models and compared their performance; a pre-trained model with the COCO dataset, and two custom-trained models with the Berkeley DeepDrive dataset, and our custom-developed dataset obtained by a Dash Cam installed onboard vehicle driven on KSA roads in five different traffic conditions; city traffic in day and night, highway traffic in day and night, and traffic in rain. We have used Google Colab platform to harness GPU power, CUDA and OpenCV. The results have been evaluated using precision and other metrics. |
|
Keywords |
Vehicle detection, tracking-by-detection, YOLO, DeepSORT, road traffic data. |
|
URL |
