This paper discussed about the video image processing technique which will helps to solve the real-time road traffic control system problems. There is a need to develop an edge detection technique with the help of real-time image tracking approach for detecting vehicles during problematic conditions and for measuring traffic queue parameters accurately.
Real-time image processing-Queue detection algorithm:
As part of image analysis, image sensors are used to capture images and ADC conversion is to convert analog video signals to digital\binary form for further processing.
Differencing through Background frame, Differencing through Inter-frame, Segmentation and classification are the three methods involved in vehicle detection. Motion detection operation and Vehicle detection operation are the two algorithms involved in Queue detection algorithm to measure traffic queue parameters. For adjusting traffic lights and to analyze cross-section traffic conditions, there is a need to measure traffic movements of vehicles at junctions. The main traffic queue parameters are the queue length, the occurrence period and the slope the queue occurrence. In order to implement the algorithm in real-time there is a need to use vehicle detection operation in a sub-profile where the queue will be extended.
Conclusions:
This paper discussed about the algorithm which helps to measure basic traffic queue parameters. Queue detection algorithm uses a technique by applying simple and effective operations. To reduce computation time, motion detection operation is applied on all sub profiles and vehicle detection operation is applied only as per requirement and it is less sensitive edge-based technique. The effects of lighting variations can be reduced by selecting the threshold dynamically. The measurement algorithm can apply to traffic scenes under different lighting conditions. Errors may cause if the objects are located far from camera and these errors can be decrease by decreasing the sub profiles size. At maximum, Queue length measurement shows 95% accuracy.
Download Image Video Processing Applied to Traffic Queue Detection Algorithm Full Report.