Optical flow based object movement tracking

Abstract

Object detection and tracking is one of the key tasks performed in video surveillance. The objects present in the area under surveillance is studied and analyzed with reference to the context. This plays a pivotal role in detecting and predicting anomalies based on the behavioral traits of objects observed under the surveillance region. Optical flow is one of the computer vision based approaches that is used for tracking the precise movement of objects. Several optical flow algorithms have been used to track and study the movement of objects. This work is motivated towards carrying out a thorough study different optical flow techniques and comparing the features of different optical flow approaches and implementing them for real time detection and tracking of objects in real time environment.