The Application of Face and People-Flow Recognition for Kanban Broadcast System
Jyun-Ciang Huang, Mao-Cheng Huang, Gwo-Jiun Horng, Gwo-Jia Jong · 2007
This paper presents face recognition combined people-flow and applied to the people-flow count in kanban broadcast system. The camera sends image data by using wireless network to computerize the face recognition for setting eigenvalue such as eyes, nose, mouth, etc. The eigenvalues are provided the data to lock the face in the program. The, principal component analysis (PCA) can be track the face using above data and find particular people in the photographed range of the camera. The face and people-flow recognition for kanban broadcast system uses database to save values for avoiding to count the same person repeatedly. An automatic feature extraction thehnique using eigen model is also demonstrated. According to the performance, we utilize the demonstration to count and analysis in the advertisement application.