Image Processing Based Automatic Pelicon Crossing System

Abstract

Traffic congestion and pedestrian accidents are two major issues that the Sri Lankan society face today. Many social, economic and environmental problems increase due to the two issues. Lack of effective Pedestrian Light Controlled (PELICON) crossing system in Sri Lanka is one of the reasons for road traffic congestion problems. Thus in this paper, we propose Image Processing Based Automatic PELICON System. The proposed system first takes an image input to the system using CCTV camera. Then, it identifies the particular color ranges using color detection algorithms. Next, noises are removed using filters. Finally, the system uses background subtraction and contour analysis algorithms to identify pedestrian object contours. The system then calculates the number of pedestrians. If the pedestrian object count is greater than a given threshold value or if the pedestrian waiting time is exceeded, then the system shows green signal to pedestrians and red signal to vehicles. Empirical study of our prototyping system has proved the effectiveness of our pedestrian detection approach. Further, we conducted a questionnaire survey to check the suitability of this system to the Sri Lankan society. Randomly selected people were taken to this study as the sample population. According to the study results, the system will minimize road traffic congestion and other negative impacts of the traffic congestion.