Real-time video stream analysis based on azure serverless services

Abstract

The object of the study is the process of obtaining additional information about the state of the domain area, which will allow performing position clarification, visual positioning, and building complex scenarios using the results of video stream analysis. The purpose of the work is to obtain additional information about the state of objects in the domain area by analyzing the video flow of agents. The proposed two-loop analysis of the video stream will allow us to quickly obtain the tagging result of each significant frame of the video stream on the hot analysis circuit, and obtain the result of object recognition on the slower cold analysis circuit. The use of data based on the video analysis of the flow of each agent allows for performing more complex scenarios, for example, target tracking ("shadow" function). Ref. 7, fig. 2.

Keywords

• IoT
• video
• frame
• stream
• analysis
• hot/cold analysis