International Journal of Engineering and Technical Research (IJETR) ISSN: 2321-0869, Volume-3, Issue-3, March 2015 297 www.erpublication.org Abstract— Millions of vehicles pass via roads and cities every day. Various economic, social and cultural factors affect growth of traffic congestion. The amount of traffic congestion has major impacts on accidents, loss of time, cost of money, delay of emergency, etc. Due to traffic congestions there is a loss in productivity from workers, people lose time, trade opportunities are lost, delivery gets delay, and thereby the costs goes on increasing. To solve these congestion problems it is better to build new facilities and infrastructure but at the same time make it smart. Many traffic light systems operate on a timing mechanism that changes the lights after a given interval. An intelligent traffic light system senses the presence or absence of vehicles and reacts accordingly. The idea behind intelligent traffic systems is that drivers will not spend unnecessary time waiting for the traffic lights to change. An intelligent traffic system detects traffic in many different ways [1]. Index Terms— Traffic density count, image processing, intelligent controlling of traffic. I. INTRODUCTION Traffic lights play an important role in traffic management. In 1868, the traffic lights only installed in London and today these have installed in most cities around the world. Sometime the vehicles on the red light side have to wait for green signal even though there is little or no traffic. It results in the loss of valuable time. Several attempts have been made to make traffic light’s sequence dynamic so that these traffic lights operate according to the current volume of the traffic. Most of them use the sensor to calculate current volume of traffic but this approach has the limitation that these techniques are based on counting of the vehicles which means it considers a small vehicle(such as motorcycle) and a big vehicle(such as truck) as the same count and so provides similar pass through time(green signal) for small as well as big vehicles. The system using image processing has been implemented where upon the density or fraction of area of road covered by vehicles is estimated and then time for green signal light is controlled accordingly. Technically, this system is based on computers and cameras. The project components includes: (A) hardware model (B) software model [2, 3]. Our project focuses primarily on the following objectives: 1. To design a system which will detect and track vehicles via camera and neglect objects which are not vehicles? Manuscript received March 23, 2015. Saima Beg, Assistant Professor Department of ECE, Integral University Lucknow, India Krishna Chandra Shukla, Post Graduation Student, Department of ECE, Integral University Lucknow, India Archana Yadav, Assistant Professor Department of ECE, Integral University Lucknow, India 2. To develop an algorithm for the above mentioned concept. 3. To develop a communication interface between the control unit and traffic signals. II. PROPOSED METHODOLOGY [Refer to figure 1]  Read video from camera or file (let the user decide the input).  Colour space converter to convert the image from RGB to intensity format.  Detect foreground using Gaussian mixture models.  Analyze frames to segment vehicles in the video.  Write the number of vehicles being tracked.  Define region of interest (ROI)  Remove the effect of sudden intensity changes due to camera's auto white balancing algorithm.  Based on dimensions, exclude objects which are not vehicles. When the ratio between the area of the blob and the area of the bounding box is above certain percentage. (to be calculated later) classify it as a vehicle.  Draw bounding rectangles around the detected vehicles.  Display the number of vehicles being tracked.  Provide User interface setting to adjust the traffic light according to number of vehicles on the road.  Also provide maximum time on green light irrespective of the number of vehicles.(To be adjustable by user). Fig.1 Image III. .PROCESSING UNIT [Refer to figure 2]  Send out DTMF signal through audio port of pc/laptop. Image Processing Based Traffic Density Estimation and Control at Intersection Saima Beg, Krishna Chandra Shukla, Archana Yadav