Adaptive Background Subtraction for Monitoring System

Security is one of the most important things for human life at this time. Complex activities often cause homes to be left unattended. One of the actions taken by most people to guard their homes while traveling is to use CCTV (Closed Circuit Television) cameras. This conventional CCTV is less effective because the camera is only recording without analyzing objects. From the shortcomings, the camera is made so that it can monitor the activities of the changes in the movement of objects seen by the camera, in this case the object detected is human movement. The monitoring system using this camera can detect passing objects. This paper proposes the adaptive background subtraction method needed to adapt to frame changes. The background frame will always be updated against the previous background intensity inference. Then it will analyze the effectiveness of the method. The effectiveness of the method used is then evaluated by comparing the results of object extraction with ground truth. The best success rate in object detection from object detection method is measured by calculating recall precision and F-measure values. The experimental results show satisfactory performance from the proposed method.