Automatic Passenger Counting System Design for Bus Using Image Processing Techniques

• Main Authors: Ting-yen Wang, 汪廷諺
• Other Authors: Jau-Woei Perng
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/34489198702212121360

碩士 === 國立中山大學 === 機械與機電工程學系研究所 === 104 === This paper to design an adaptive passenger counting system is applied to the public transportation vehicle: bus. The system combines Global Positioning System (GPS) and standard time to establish number of passengers for each bus stop in different period time and calculates the number of the passengers and unoccupied seats to provide Bus service company business-related information. People counting is one of many important applications of video surveillance systems. In this paper, we propose an automatic people counting system by applying background subtraction and object tracking on video camera output. The foreground is extracted by background subtraction. Morphological operations are used to repair damage to the image and connected-component labeling method is used to eliminate of noise. Background subtraction identifies moving objects in the form of connected components. The people detection algorithm attempts to determine whether moving objects correspond to the human model that established by back propagation neural network. K-means clustering is used to enable the segmentation of single persons from connected-component foreground. Each centroid of foreground objects is used to track moving objects for counting.