Automatic brake Driver Assistance System based on deep learning and fuzzy logic
2024

Automatic Brake System Based on Traffic Lights

publication Evidence: moderate

Author Information

Author(s): García-Escalante A. R., Fuentes-Aguilar R. Q., Palma-Zubia A., Morales-Vargas E.

Primary Institution: Tecnológico de Monterrey

Hypothesis

Can an automatic brake system effectively detect traffic lights and apply brakes based on their states?

Conclusion

The proposed system achieved a 96% accuracy in detecting traffic light states and a 1 meter average error in distance estimation during real-world tests.

Supporting Evidence

  • The traffic light detection model achieved a mean Average Precision (mAP) of 0.96 for distances less than 13 m.
  • The system's response time was measured at 0.23 seconds.
  • An average Root Mean Squared Error (RMSE) of 1 m was found for distance estimation during on-road experiments.

Takeaway

This study created a smart braking system for cars that can see traffic lights and stop the car when needed, making driving safer.

Methodology

The study used a Jetson TX2 and a ZED camera to detect traffic lights and apply brakes based on their states using fuzzy logic.

Limitations

The system struggled to accurately estimate distances to traffic lights beyond 13 meters due to environmental factors.

Digital Object Identifier (DOI)

10.1371/journal.pone.0308858

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication