Within SETO project, Cerema is advancing sustainable infrastructure management through AI, by developing a method using convolutional neural networks named YOLO (You Only Look Once) to detect lifted truck axles from camera images, improving truck classification and traffic management.
During the latest applications of this methods, confidence score has reached 98%, as highlighted in a recent article published on Cerema’s website: https://lnkd.in/d5HC6Gqz
Next Steps will include:
- Expand training with diverse data.
- Enable real-time processing for high-traffic scenarios.
- Implement distance measurement between axles.
These findings will be also presented at HVTT Forum HVTT18 Conference that will take place in 2025 in Quebec, Canada.
