Docker container based analytics at IoT edge Video analytics usecase

The number of internet connected devices is growing rapidly. It is getting difficult to gather and process the huge volume of the data generated from these devices. Edge computing paradigm is evolving that helps to process the data locally close to the source. This paper highlights using docker container based analytics services at the edge for data processing. Feasibility is studied for setting up deep learning framework on Raspberry Pi for real-time analysis of video feed from the surveillance camera. Performance benchmarking is also done to figure out overhead of containerization.