Volume 15, Issue 5

AI-Driven Adaptive Traffic Management System Using YOLOv8 for Real-Time Congestion Optimization

Author

Suhani Prakash, Vaishnavi Raghuvanshi, Yash Kumar Gautam

Abstract

The article describes an adaptive traffic control system based on artificial intelligence to optimise traffic. The new system is an adaptive one, as compared to the traditional fixed-time traffic light system, which has a fixed cycle and can cause congestion. Using computer vision and deep learning, such as the YOLOv8 object detector, vehicles are detected and classified from CCTV video (cars, bikes, buses, trucks, etc). The traffic density is calculated using a weight value for different types of vehicles in each lane. The traffic light timings are then optimized to minimize traffic congestion. Ok, let's try to make it sound a little more like humans. To test the performance of the system, the system is tested with the Webster delay model, to measure, among other things, the average delay time. The tests are compared to the conventional fixed-timing systems and it's evident that traffic congestion is reduced and traffic flows smoothly. There's also a minor improvement in fuel efficiency, so this traffic control could be more than just an idea.

REFERENCES

[1] Kumar et al., "Vehicle detection using image processing techniques," International Journal of Computer Vision Applications, 2019.

[2] Gupta and Singh, "CNN-based deep learning for vehicle classification," International Journal of Artificial Intelligence Research, 2020.

[3] Patel et al., "Traffic density estimation using YOLOv3," IEEE Conference on Intelligent Transportation Systems, 2021.

[4] Rao and Nair, "Machine learning-based adaptive signal timing," IEEE Transactions on Intelligent Transportation Systems, 2022.

[5] Mukherjee et al., "Real-time traffic density estimation using YOLOv5 and ML prediction," Journal of Smart Transportation Systems, 2023.

[6] Ali et al., "Vehicle detection using background subtraction method," International Journal of Image Processing, 2019.

[7] Verma et al., "Vehicle detection using Faster R-CNN in dense traffic," International Conference on Computer Vision Systems, 2020.

[8] Singh and Reddy, "Traffic flow analysis using YOLOv4," International Journal of Advanced Computer Science, 2021.

[9] Chavan et al., "Multi-object tracking using DeepSORT and YOLO," IEEE Access, 2022.

[10] Banerjee et al., "Traffic prediction using LSTM networks," Journal of Transportation Engineering, 2021.

[11] Ahmed and Khan, "SVM-based vehicle classification for traffic systems," International Journal of Machine Learning Applications, 2020.

[12] Bhosale et al., "Real-time vehicle detection and tracking using YOLOv5 and SORT," IEEE International Conference on Smart Cities, 2022.

[13] Li et al., "YOLOv8-based real-time vehicle detection system," IEEE Access, 2023.

[14] Kim et al., "Edge AI-based traffic control for smart cities," IEEE Internet of Things Journal, 2024.

DOI

https://doi.org/10.62226/ijarst20262670

PAGES : 1933-1938 | 10 VIEWS | 3 DOWNLOADS

Download Full Article