On a Distributed Video Surveillance System to Track Persons in Camera Networks

Abstract

Contribution of the ThesisIn automated video surveillance applications, presenting the useful information to the human operators is a challenging task. Current systems usually require a prohibitive amount of human resources and lead to a quick decrease of the attention of the human operators through time, thus preventing them to catch the relevant events that may be worth to further investigate.In addition, when monitoring a wide area, it becomes hard to deploy a network of video sensors such that there are enough overlapping FoVs to cover every point of the environment. This leads to the development of video surveillance systems (VSS) that provide partial area coverage. As a result, "blind-gaps" between camera FoVs are introduced. One of the most interesting problems which such "blind-gaps" bring in is to re-identify the persons moving across disjoint FoVs.The contribution of the thesis * is two-fold. First, an advanced VSS is designed to display the proper taskdependent information to operators that are monitoring a wide area. In particular, the system helps operators in the task of tracking persons across camera views. This raised the need for a system capable of re-identify the subjects moving through disjoint FoVs. This leads to the second contribution of the thesis, a distributed approach to address the challenges of the person re-identification problem. Adaptive Human Interface for VSSThe main goal of the proposed VSS is the development of an effective information visualization technique that properly displays only the most relevant cameras and information contents to simplify the operators' tracking tasks [1,2]. The key idea is to visualize only most probable streams, i.e. those that will be involved with the motion of the tracked persons. Towards this goal we propose a novel dynamic organization, activation and switching of the User Interface (UI) elements based on the output of video analytics algorithms.The first objective is to distill the volumes of monitoring information into a human manageable quantity. This is achieved by introducing an hand-off task between different camera views so that a single person can be tracked across different FoVs. The proposed camera planning algorithm uses geographical clues and exploits the predicted trajectories to build an accurate camera activation plan.