JIITA, vol. 2, no. 3, pp.136-139, Sep, 2018
Real-time People Counting System Using Simplified Motion Detection
Sung In Cho 1,*), Suk-Ju Kang 2)
1) Department of Electronic Engineering, Daegu University, South Korea.
2) Department of Engineering, Sogang University, Seoul, South Korea.
Abstract: This paper proposes a real-time system for estimating the count of pedestrians. The proposed method consists of two main procedures: estimation of binary motion and motion flow rate on a line of interest (LOI). By modeling a background, the motion generated by a human movement can be simply extracted. Then, the motion flow which is the number of motion pixels crossing the LOI is calculated. Finally, the number of people crossing the LOI is generated by normalizing the motion flow. Simulations show that the proposed method successfully estimates the number people crossing LOI with simple operations.
Keyword: component; formatting; style; styling; insert (key words)