Traffic data is an important factor used to determine the signal timings

Traffic data is an important factor used to determine the signal timings. Detectors are being used in traffic signal system for a long time. Most of the earlier traffic signal systems used ultrasonic sensors, microwave sensors, inductive loop and infrared sensors. But they had few disadvantages, such as, poor performance in noisy environment, unable to detect motionless or slow moving objects and pedestrians. With improvement in detections technologies like radar, video detectors, global positioning system, and radio frequency identification, modern traffic signal systems are more reliable. Although, most of the above mentioned technologies are used in traffic control system, video surveillance is one of the most widely used detectors to monitor traffic in adaptive signal controllers. Advantages of using video detection techniques in traffic management systems are:
• Ability to detection motionless and pedestrians.
• Allows for data collection and flow monitoring.
• Highly reliable at intersections.
• Ability to detect incidents quickly.
In this section, we will discuss more about implementation of video detectors.

Intelligent video analytics allow traffic controllers to effectively manage the traffic flow. A camera sends an input signal to a detector when a vehicle enters a detection zone. The detector captures the image and based on the pixel difference, the detection is activated. The analyser generates various types of traffic information and the data is sent to controller. Based on the monitored data, the controller takes appropriate decisions.

1 Designed vision surveillance with adaptive features in order to measure traffic on a busy road. The proposed model filters motionless and slow moving objects on iterative comparisons with previous images. It involves three processes:
• Average Method: computes the average value of pixel for current and previous image.
• Inpaint Method: comparing two continuous images and finding the difference between their pixel’s lightness to check whether it belongs to background image or not. If the difference between them is higher than a specified value, an image processing algorithm is used to find the number of vehicles by comparing it with background image.
• Building Background steps: If the pixel difference between two continuous images is less than the specified value, it stores the image as background and is used in next comparison.
An adaptive traffic control system is built using grid agent technology. Controller receives data from neighbouring detectors, based on the monitored data a decision is made on signal timings. The experimental results show reduction in noise and unwanted disruptions in monitored data. 2 Proposed a hybrid video detector based on edge detection algorithm for intelligent traffic controllers. The process of the proposed model is similar to 1, where, the current images are compared with background image. The model was tested in real conditions and the accuracy of detector was above 95% for both vehicles and pedestrians.