Automatic vehicle detection and motion path tracking based on gaussian mixture model

Abstract

Automated traffic surveillance plays a vital role to build up a smart transport and communication system of any modern city. This paper deals with three primary aspects of any electronic traffic control system viz., identifying moving vehicle intruding a specified traffic zone, its count and tracking of the motion path. Gaussian mixture model (GMM) is used to detect the foreground object followed by blob analysis which in turn gives the moving vehicle identification. Finally a new concept of weighted mean is deployed for motion path tracking.