Autonomous Passenger Counting System with Distributed IP Cameras – Phase 2

The City of Edmonton is currently in need of an automatic passenger counting (APC) system for its Light Rail Transit (LRT). An APC will help the City with various decisions making. In this proposal we focus on automatic video processing to count passengers boarding and alighting from LRT. The proposed system is simple – it will consist of overhead inexpensive IP cameras directly looking down at the car doors. Direct top view will be able to avoid occlusions. Low-to-moderate resolution, grayscale video frames will be transferred over a network to a processing center and processed there automatically to count passengers. The goal of this four-month internship is to perform a feasibility study of the proposed setup and provide further recommendation to the City of Edmonton toward implementing a future APC.

Faculty Supervisor:

Dr. Nilanjan Ray

Student:

Satarupa Mukherjee

Partner:

AQL Management Consulting Inc.

Discipline:

Computer science

Sector:

Construction and infrastructure

University:

University of Alberta