Abstract:
As there is sudden increase of traffic on road due to the utilization of personal cars. This sudden increase in traffic results in traffic congestion because traffic signal is working without any vehicle density calculations. The proposed system work on image processing techniques to solve this problem. It performs the calculation of vehicles during night time to overcome the traffic congestion. System performs the identification of vehicles and in this way control the traffic density. Objects are identified and counted using some image pre-processing and blob detection techniques. This suggests the requirement of an intelligent video closed-circuit television providing continuous 24–hour observation, replacing the standard ineffective systems. The proposed work can lead to an efficient signalling system and provides a basis of real time implementation of pc vision technology in real time. Moreover, the proposed work introduces new schemes to increase the efficiency of the traffic signalling system.
This paper presents an automatic vision primarily based traffic management system that is capable to observe and track vehicles from a video. Simulation results have shown that the article Classification module manages to attain associate accuracy of 91.14% for vehicle detection. Moreover, the system manages to successfully track the objects 91% of the time below no occlusion and 91.14% in presence of occlusion.
