Traffic Surveillance system is being more important with the increasing number of vehicles. Much better ways for
traffic analysis are also developed. Traffic analysis is the analyzing the cloud of vehicles in defined place for specific interval of
time and the vehicle classes. Now a day most of the people involving in sensor use to detect the vehicles. Even though these
systems are highly effective and matured and not very budget friendly. These systems need high maintenance and periodic
calibration; so it causes highly smarter computer vision based systems for traffic surveillance. Our proposed system takes the
input data in the form of RGB images and converts into gray level images. System extracts the features of different vehicles and
gives a vehicle counter and classifier based on combination of several different video-image processing techniques including
background Subtraction, Gaussian mixture model(GMM), frame differentiation, Gaussian blur filter.
