Towards an Autonomic Intelligent Video Surveillance System
Abhishek Djeachandrane, Said Hoceini, Serge Delmas, Abdelhamid Mellouk · 2025
This chapter presents a novel autonomic approach for an Intelligent Video Surveillance (IVS) system. Three decision support systems (DSS) are proposed based on machine learning (ML), descriptive statistics and inferential statistics. The first system, User-centric self-configuring DSS, does not require an event trigger and explains how to deploy an autonomic IVS for self-configuration. The second system, User-centric self-protecting DSS, triggers an event when the video quality is not suitable. This is done using descriptive and inferential statistics, which draw some conclusions about the samples and allow the system to self-protect from noise injection. The third system, User-centric self-awareness and self-learning DSS, triggers an event when the video quality is not suitable to learn, and we will enable the ML algorithm to be aware of the situation on its own to trigger its learning.