การวิเคราะห์คุณลักษณะการเดินบริเวณโถงทางเดินภายในสถานีรถไฟฟ้าเตาปูนจากกล้องโทรทัศน์วงจรปิด

Abstract

Public transportation is widely recognized as a sustainable mode of travel with minimal environmental impact. Currently, the railway system is widely regarded as the fundamental infrastructure supporting public transportation in Bangkok. The implementation of such a system necessitates its deployment in a database that may not align with the preferences and requirements of Thai users. The primary aim of this study is to discern the attributes of pedestrians and modify the variables pertaining to the flow of pedestrian traffic. The objective is to collect pertinent data to facilitate the design of pedestrian walkways within the confines of a train station. The procedure entails the utilization of Closed-Circuit Television (CCTV) technology for the purpose of capturing video footage within Tao Poon MRT Station. The objective is to create computer vision software utilizing the Python programming language, with the implementation of the YOLO model for the purpose of person detection. The utilization of a fundamental variable correlation model is employed to modify the parameters in response to varying pedestrian traffic flows. We investigate the correlation between fundamental variables associated with pedestrian traffic flow. The parameters are calibrated utilizing data obtained from the study area, which unveiled an average velocity of 82.93 meters per minute for walking. The results indicated that Underwood's model exhibited the greatest level of accuracy and consistency in the study area. The study suggests that the maximum rate at which pedestrians flow is 149 persons per meter per minute. Furthermore, the findings of the study indicate that pedestrians have a maximum attainable walking speed of 86.34 meters per minute. The investigation reveals that the optimal flow rate is attained when pedestrians maintain a walking speed of 32.50 meters per minute.