Image processing system for intelligent transportation system

Abstract

Monitoring heavy road traffic to tackle the issue of road congestions has always been an important concern for road authorities. To pave the way for countering this prominent issue of road congestions, traffic analysis systems used must first be high in efficiency and accuracy. By obtaining road traffic statistics which are highly accurate, road authorities can then implement the necessary steps to optimise road traffic conditions. In this study, current traffic data monitoring methods were investigated. Different image segmentation techniques and image enhancement methods will be explained and tested out on traffic video footages. Frames of traffic video footage samples were taken and extracted into the required number of frames using the MATLAB interface. Using the GUI (Graphic User Interface) programme, comparative analysis of different combinations of image segmentation methods and image enhancement methods was carried out to find out which produced the most accurate results. On top of that, the execution of time of the different image processing techniques were also recorded and compared in this report. Factors which could affect the accuracy of the results were discussed, as well as the recommendations for future work.