Theme - GIS for effective Traffic management of Bangalore (Special Category)
The traffic situation in Bengaluru is inevitably a matter of trouble for its commuters. A city with such thriving population needs state-of-the-art Technology to attend to its high infrastructure needs. This project "AI for Traffic Signal Control" aims to maximize the tech utilization to come up with a frugal and adaptive solution for the problem.
Preliminary Objectives:
- 40% reduction in average waiting time
- Pre-emption system to ease movement of priority vehicles (Ambulances, Fire engines)
The proposed system aims to minimize waiting times for motorized vehicles in traffic junctions. The system extracts data like Queue Lengths and Average Speeds of vehicles from live video feed. A Deep Q-Learning reinforcement model optimizes traffic light configuration based on extracted data.
An additional pre-emption system is planned to ease priority vehicle movement. Crisis regions are mapped for all junctions across the city. Location data of priority vehicles is used to activate green lanes for junctions.
NOTE - Descriptions and Instructions to Run are in the Readme within Modules
Extraction of vehicular density data from video feed
- OpenCV
- FCNN Algorithm, YOLO Architecture
- Simuation of Deep Q Learning
- Sumo(Simulation of Urban MObility) tool with TraCI module for traffic interfacing
- Pitch - https://youtu.be/JMx4zSGA54U
- Demo - https://youtu.be/QPFXY4AR5Pg