Code assets for BlueGranite's "Edge-Based Passive Street Monitoring for Smart Cities" solution from the Microsoft AI: Look, Listen, Innovate! Solution Hack event in March 2021.
In large urban settings, the monitoring of streets is cumbersome task for municipalities to manage. This is important information to have as short- and long-term decisions are made at the street block level including maintenance scheduling, transportation planning, and understanding crowd flow. Today, the collection of this data is limited and often relies on humans collecting this data manually over time.
Currently, the use of surveillance cameras can also be used to remotely monitor these conditions, however, these cameras can raise questions around public privacy as to what information is being captured and how it is being used.
Understanding multi-model transportation in cities allows for better planning, more efficient use of resources, and provides better insights into public behavior. The use of computer vision-based monitoring of streets will provide deeper insights into pedestrian density, parking scarcity, and road conditions.
Our edge-based approach ensures the privacy of public data while providing real-time information to municipalities across their city. This AI-centric monitoring solution will be able to remotely track multiple factors on a given city street.
Our solution is to use AI models deployed on the edge for quickly capturing the bits of information needed for city street monitoring without saving huge amounts of data or concerning the public with privacy issues.
This innovative approach will blends the power of the Azure Cognitive Services with custom trained models to provide insights in a performant, unified platform for street monitoring. By scoring incoming video data from a small camera device on the edge and relaying the output data back to the cloud through IoT Hub, we can enable real-time insights across a city.
- Project Santa Cruz GitHub: https://github.com/microsoft/Project-Santa-Cruz-Preview
- People Detector App Template: https://github.com/microsoft/Azure-Percept-Reference-Solutions/tree/main/people-detection-app