Study and Implementation of an Object-based Video Pre-encoder for Embedded Wireless Video Surveillance Systems

Résumé

Embedded wireless video surveillance systems are gaining widespread popularity due to advancements in internet of things (IoT) systems and wireless multimedia sensor networks. These systems have numerous applications, including military target tracking and surveillance, disaster relief, biomedical health monitoring, seismic sensing, environment monitoring, and smart cities. Transmitting multimedia data at a low bitrate while maintaining high-quality transmitted data is still a challenging problem in battery-powered systems due to limited energy availability. This thesis addresses this challenge by optimizing energy efficiency in wireless multimedia sensor networks for resource-constrained wireless surveillance environments. The focus is on developing novel encoders that minimize energy consumption while maintaining high Quality of Experience (QoE) for both human and machine processing. The thesis introduces low-complexity methods for detecting regions-of-interest (ROI) in video frames. This will enhance accuracy and robustness by leveraging multiple object detection techniques. These detection techniques are integrated as pre-encoders in different encoding chains for wireless video surveillance, resulting in significant energy and bitrate savings of up to 98% while preserving acceptable quality of service (QoS) and QoE. Several tests and experiments demonstrate the feasibility and effectiveness of the proposed approaches in this thesis. The findings of this research pave the way for future research in this field

Citer ce document

Accès au document

Ce lien n'est plus accessible actuellement. Contactez l'institution d'origine.

Auteur(s)

Statistiques

Consultations : 4

Téléchargements : 0