A Benchmark of Motion Detection Algorithms for Static Camera: Application on CDnet 2012 Dataset

Abstract

The main aim behind this study relates to comparing a variety of motion detection methods for a mono static camera and it endeavors to identify the best method for different environments and complex backgrounds whether in indoor or outdoor scenes. For this reason, we used the CDnet 2012 video dataset as a benchmark that comprises numerous challenging problems, ranging from basic simple scenes to complex scenes affected by shadows and dynamic backgrounds. Eleven ranging from simple to complex motion detection methods are tested and several performance metrics are used to achieve a precise evaluation of the results. This study sets its objective to enable any user to identify the best method that suits his/her need.