Intelligent Pedestrian Signal Light System
Final Year Research Project • Published on ResearchGate
Overview
An intelligent traffic management system that adapts signal timing based on real-time pedestrian detection, weather conditions, and traffic flow to improve safety and efficiency.
Role
Lead Developer & Researcher
Timeline
2023 – 2024
Key Metrics
80%
Pedestrian Detection Accuracy
18.5s
Average Crossing Time (reduced from ~60s)
5–10×
Faster Real-time Processing (vs Arduino)
Tech Stack
Challenges
Real-time processing on embedded hardware, accurate detection in varying weather/lighting.
Solution
Custom CNN model trained on labeled dataset • Upgraded to Raspberry Pi 4 for multi-stream handling • IoT framework for sensor-camera-controller communication • Adaptive timing algorithm.
Outcome
Deployed prototype reduced crossing time in real-world case study • Violation deterrence via sirens/alerts • Full publication available on ResearchGate.





