Image processing system for qualitative road traffic data analysis

Abstract

Road traffic data analysis system has been a main concern of traffic engineer to optimize the road traffic and reduce the vehicle incident rate. This project aims to develop a high efficiency and high accuracy traffic analysis system, so that it can be used in real time road traffic detection system to monitor traffic flow and collect statistical data to further optimize the road traffic condition. The comparative data analysis of several image processing techniques is illustrated in this report. The image processing techniques include background difference, inter-frame difference, binary conversion, edge detectors and quadtree decomposition methods. Upon the data analysis, an optimal image processing techniques can be applied to detect vehicle for further implementation and analysis. Different image processing techniques have different characteristic, advantages, and disadvantages. Road traffic sample was recorded and before undergoing the image processing techniques in the system, the video was extracted into frames. The comparative analysis was carried out based on several brightness adjustments. In addition, the execution time and accuracy represent a system efficiency especially in real-time image processing system. A comparison result on the execution time of different image processing techniques is also presented in this report.