Detection of moving objects: The first stage of an autonomous surveillance system

Abstract

Object detection is an essential first stage in a surveillance system, primarily because it focuses all the subsequent processes. The standard approach to object detection is background subtraction. At the core of background subtraction is a module that maintains an image that is representative of the scene monitored by a camera. This work compares two background subbtraction/maintenance algorithms: adaptive Gaussian mixture model and the Wallflower method. The algorithms are evaluated using video footage of the real world. The Receiver Operating Characteristic (ROC) curves are used to quantify the performance of the algorithms. In our experiments, the adaptive Gaussian mixture model outperforms the Wallflower method.