A real time passenger counting by using simulink

Abstract

This project presents a passenger counting system which counts and tracks moving people in a video sequence. The system works well under a stationary background with changing illumination. The statistics achieved from the counting system are useful in population flow. For example, the video sequence recorded at the entrance of a building is used to evaluate the total number of visitors at a specific time. In this counting system, background estimation technique is applied. By subtracting the background from each video frame, moving objects (walking people) are extracted to produce the foreground image. By threshold and performing morphological closing on each foreground image, binary feature images are produced. Finally, blob analysis is used to compute the bounding box of each person in order to count the number of people.