Skip to content

mpignatelli12/PedestrianBuddies

Repository files navigation

Pedestrian Buddies

Pedestrian Buddies is an embedded computer vision project that integrates traffic signals with crossing-guard gesture recognition.
The goal is to make intersections safer by automatically interpreting a guard’s hand signals and controlling lights in real time.


Features

  • Real-Time Gesture Recognition (~30 FPS on edge hardware).
  • Traffic Signal Integration — triggers stop/go based on detected gestures.
  • Robust Model — tuned for variable lighting and occlusion.
  • Edge Deployment — optimized for Raspberry Pi / microcontroller-class devices.

Tech Stack

  • TensorFlow Lite, Keras
  • Python
  • OpenCV
  • Arduino (or equivalent edge hardware)

About

Integrating traffic signals with a computer vision system to detect crossing guard gestures. Built with TensorFlow and TensorFlow Lite.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors