| This thesis presents a methodology for extracting traffic parameter using video image processing. It is an offline system that focuses on extracting vehicles, tracking them and provides the vehicle count for a short period of time. It uses background subtraction algorithm for extracting moving objects. The salient feature of this research is that it uses a novel method for extracting vehicle count. The proposed method for counting vehicles uses screen-line that keeps track of each vehicle crossing it. The results show that the algorithm is capable of counting 100% of the vehicles. Some noise was induced during the cast shadow removal process. Due to the presence of such noise pixels, final vehicle counts were increased by approximately 15%. |
