Comparative Review on Video Based Vehicular Traffic Data Collection for Intelligent Transport System

The compilation of high quality traffic data is one of the basic concerns for real time traffic video surveillance. Since Video surveillance has figurative applications such as security, traffic monitoring and law enforcement. Detecting and recognizing the vehicle objects in a video is an imperative part of video surveillance systems, which is going to guide in tracking of the detected objects and gathering decisive information. Vehicular detection and classification is decisive as it can guide in traffic control and gathering of traffic info that can be used in intelligent transportation systems. Vehicular classification poses a thought-provoking problem as vehicles have high intra-class variation and comparatively low inter-class variation and it is time consuming, varying performance under different weather conditions etc. Hence there is a demand for employing innovative efficient data collection technique which will give the advantageous continuous data collection with acceptable level of accuracy. This paper surveys the prominent published literature of last few years to evaluate and summarize the work done so far on this field. The motivation for this paper is to provide a review of current methods employed for traffic data collection for Intelligent Transport System (ITS) available in the literature and estimate best one that accomplish all parameter conditions.