Writing your client application component - Amazon Lookout for Vision

End of support notice: On October 31, 2025, AWS will discontinue support for Amazon Lookout for Vision. After October 31, 2025, you will no longer be able to access the Lookout for Vision console or Lookout for Vision resources. For more information, visit this blog post.

Writing your client application component

A client application component is a custom AWS IoT Greengrass Version 2 component that you write. It implements the business logic you need to use an Amazon Lookout for Vision model on an AWS IoT Greengrass Version 2 core device.

To access a model, your client application component uses the Lookout for Vision Edge Agent component. The Lookout for Vision Edge Agent component provides an API that you use to analyze images with a model and manage the models on a core device.

The Lookout for Vision Edge Agent API is implemented using gRPC, which is a protocol for making remote procedure calls. For more information, see gRPC. To write your code, you can use any language supported by gRPC. We provide example Python code. For more information, see Using a model in your client application component.

Note

The Lookout for Vision Edge Agent component is a dependency of the model component that you deploy. It is automatically deployed to the core device when you deploy the model component to the core device.

To write a client application component, you do the following.

  1. Set up your environment to use gRPC and install third-party libraries.

  2. Write code to use the model.

  3. Deploy the code as a custom component to the core device.

For an example client application component that shows how to perform anomaly detection in a custom GStreamer pipeline, see https://github.com/awslabs/aws-greengrass-labs-lookoutvision-gstreamer.