Visualize AI events in real-time from Azure Video Analyzer using Power BI
Published Sep 17 2021 09:00 AM 3,463 Views

Ever wondered how you could improve your business dashboards with real-time visualization of AI inference events combined with a video playback experience? 


Globally cities are becoming smarter by leveraging cameras and real time analytics to manage highway congestion. Identifying traffic peaks locally, understanding these peaks across the entire city and managing the traffic routing in real time needs scalable video analytics as well as intuitive and aesthetically pleasing visualization infrastructure. This is where the power of Azure Video Analyzer with Microsoft Power BI comes to the fore.  


In this blog we will use Video Analyzer's new capability called Line Crossing which allows you to define a virtual line on a sample highway video and then count the number of vehicles crossing this virtual line. Using Video Analyzer, you can publish the results of this analysis to IoT Edge Hub in the form of AI inference events. These AI inferences can then be routed to Power BI via Azure Stream Analytics to gain insights and share dashboards with peers in your organization. 


Overview of steps to create dashboard from AI events in Power BI:

  1. Create and run a Stream Analytics job to retrieve necessary data from IoT Hub and send to Power BI
  2. Run a live pipeline within Video Analyzer that generates AI inference events and pushes them to the IoT Hub
  3. Create a Power BI dashboard to visualize the AI inference events
  4. Optionally embed the Video Analyzer player widget in the same dashboard


Detailed instructions:

  • Step-by-step instructions in the form of a complete tutorial can be found here. 
  • Step-by-step instructions to embed player widget can be found here. 

Follow the two articles to create a dashboard as shown in the screenshot below - 




Version history
Last update:
‎Sep 17 2021 09:14 AM
Updated by: