Consider the scenario of detecting the void spaces on the shelves of a retail store. Your goal is to run machine learning models on your cameras, train models for your objects, and deploy at the edge devices so that voids on shelves can be detected. Live Video Analytics(LVA) on IoT Edge enables you to utilize your existing cameras to generate video feeds. Using Custom Vision, you can manually capture your images from existing camera feeds, label your objects and deploy the model so built at the edge.

You perform analysis at the edge and then publish the results to other services such as business logic or directly to Azure Services in the cloud. So this combination enables you to build business applications that are spanning the edge and the cloud.

Clean up

When you're working on your subscription, it's a good idea to identify whether you still need the resources you created at the end of a project. Resources left running can cost you money. You can delete resources individually or delete the resource group to delete the entire set of resources.