Publication

Abstract : Security threats have always been a primary concern all over the world. The basic need for surveillance is to track or detect objects of interest over a scene. In most of the fields, computers are replacing humans. One such field where computers play a great role is surveillance. Typically, computer based surveillance is achieved by computer vision that replicates human vision. Here, sequences of panoramic images are created and moving objects are detected over a particular time. Moving objects are being detected using various background subtraction methods like Frame Differencing, Approximate Median and Mixture of Gaussians. In real-time applications like surveillance, the time that takes to make a decision is critical. Hence, a comparison is made between these methods in terms of elapsed time. © 2018, Springer Nature Singapore Pte Ltd.