Communication Protocol Between Embedded Computers and Fog Computing Environment for Image Processing

In this article a new data transmission protocol between embedded computers and the fog computing environment for image processing is considered. The possibilities of using fog environment for intelligent processing of video are revealed. The problems of transmission protocols are indicated. A comparison of existing video processing systems is given. The place of embedded computers in intelligent video surveillance systems and ways of upgrading these systems are considered. Data from the devices is collected using embedded computers and visualized using IoT technologies. The developed data transmission protocol allows processing images in the fog and/or cloud for intelligent video surveillance systems. It assumes packet data transmission of series of frames with additional information that are processed using machine learning algorithms and neural networks. The effectiveness of the new systems is shown. As an example, the problem of processing video coming from cameras used in the subway to determine damage of the escalator tape and steps is considered. Systems for image processing that implement the proposed protocol can be used not only in the subway, but also in many other areas where Internet of things technologies are supported.