Abstract
The information of visitor that entering in an area could be an important
information for several purposes such as business, security, resources demand,
and etc. While numbers of visitors are too excessive that could reduce accuration
of manual counting, then this condition needs a system which could count
automatically and people counting could do it.
The developing of computer vision could give good accuration when
implementing it in people counting. Using Face Detection wih haar-cascade
classifier that can give accuration 94%, tracking with extended CAMSHIFT can
avoid tracking window collision with the same-color object with accuration 88%.
This method can be implemented to people counting.
Keyword : Computer Vision, People Counting, Haar-Cascade Classifier,
Extended CAMSHIFT