Title: Novel context-aware methodology for risk assessment in intelligent video-surveillance systems
Authors: Isaac Martín de Diego; Alberto Fernández-Isabel; Ignacio San Román; Cristina Conde; Enrique Cabello
Addresses: Data Science Laboratory, Rey Juan Carlos University, C/Tulipán, s/n, Móstoles, Madrid, Spain ' Data Science Laboratory, Rey Juan Carlos University, C/Tulipán, s/n, Móstoles, Madrid, Spain ' Data Science Laboratory, Rey Juan Carlos University, C/Tulipán, s/n, Móstoles, Madrid, Spain ' Face Recognition and Artificial Vision Group, Rey Juan Carlos University, C/Tulipán, s/n, Móstoles, Madrid, Spain ' Face Recognition and Artificial Vision Group, Rey Juan Carlos University, C/Tulipán, s/n, Móstoles, Madrid, Spain
Abstract: Nowadays, the presence of multiple cameras is steadily increasing. Therefore, video-vigilance systems need to be massively produced. These systems are controlled by human experts who spend excessive amount of time detecting problems or possible impediments. Intelligent video-surveillance systems appear to mitigate this issue. Nevertheless, these systems present their drawbacks. One of the most important drawbacks is their inability to process the obtained visual information and use it to modify their parameters according to the environmental requirements. In this paper, a novel methodology has been developed to provide information to a previously implemented intelligent video-surveillance system becoming sensible to the context. The methodology is based on the advice of experts to adapt the system for a given scenario. It includes both the information from generated alarms and knowledge about the environment. The experimental results have been promising and the proposed methodology can serve as the foundation for future enhancements.
Keywords: risk assessment; intelligent video-surveillance system; context-aware methodology; fuzzy logic; internet of things.
DOI: 10.1504/IJSNET.2022.127121
International Journal of Sensor Networks, 2022 Vol.40 No.3, pp.145 - 159
Received: 22 Jan 2022
Accepted: 25 Jan 2022
Published online: 22 Nov 2022 *