Intelligent transportation system based on image processing

In today’s global digital world, traffic congestion is becoming a serious problem. Previous studies have provided several systems which are being used to handle traffic problems. However, each of them has its own disadvantages. In contrast, a system based on image processing is better for users. It is because this system uses image processing algorithms to get accurate data, which is more convenient and costs less to be implemented. The main objective of this project is to design and develop an intelligent transportation system based on image processing to extract necessary traffic data for better road management. The data includes vehicle count, vehicle types, vehicle speed, road usage and traffic accidents. Another objective of this project is to do the data analysis, to compare the performance of image processing methods in different environmental conditions. All the image segmentation methods and road traffic analysis algorithms have been realized by MATLAB functions. In this project, the performance was judged by the error rate of counting the total number of vehicles for each lane. In normal condition and rain condition, most algorithms yielded satisfactory results, whereas in rush hour condition and night condition, the accuracy of most algorithms declined. In addition, in general, edge detection methods and quadtree decomposition are more reliable than background subtraction methods and binary image conversion. The performance of most algorithms is as expected. The author has made possible explanations of some results and given possible solutions. Recommendation for future work is also proposed for areas of improvement.Bachelor of Engineering (Electrical and Electronic Engineering