Task Offloading Strategies for Resource-Constrained Edge and Fog Layers in Object Recognition

Seeing the fast growth of Internet of Things (IoT) devices has increased data generation, making job management and processing more vital than ever. Traditional cloud-based solutions might impede real-time applications due to latency and bandwidth issues. The research presents a method for energy-efficient Edge-Fog-Cloud computing image processing and workload offloading method. Resource optimization, energy conservation during task loading and real-time processing performance are achieved through dynamic offloading between edge fog and cloud by using deep learning-powered object identification in videos. The method uses YOLOv5 module for accurate object identification and Azure IoT Edge for deployment on cloud. By reviewing various methods for optimizing use of cloud services in IOT architecture it can consider methods like Energy-Efficient Fog Computing Framework to improve workload distribution and power management. The technology balances computational efficiency and energy savings with edge, fog, and cloud 3 layers architectures. By developing efficient as well as scalable image processing in video based on the Internet of Things, this study advances Fog-Cloud computing for video surveillance. Intelligent task offloading and adaptive processing optimize system performance and responsiveness in real-time, judging from the results. Further research will improve these solutions for IoT application flexibility.