A person using a self-checkout kiosk

Automated Self-Checkout Retail Reference Implementation

Automated Self-Checkout Retail Reference Implementation provides:

  • High-quality testable ingredients needed to create an automated self-checkout solution.

  • Building blocks needed to enable the use case with solid performance numbers.

  • Hardware requirements to support the workloads and scale the solution.

  • Predictable return on investment (ROI) and lower total cost of ownership (TCO) to deploy and manage the solution.

author-image

作者

  

Overview

The Automated Self-Checkout Reference Implementation provides critical components required to build and deploy a self-checkout use case using Intel® hardware, software, and other open source software. This reference implementation provides a pre-configured automated self-checkout pipeline optimized for Intel® hardware.

Requirements 

To build the Intel® Automated Self-Checkout Reference Implementation, you need:

  • Ubuntu* LTS Boot Device
  • Docker*
  • Git*

To know about the supported platforms, see the list of platforms.

Learning Objectives

Using this reference implementation, you can:

  • Identify optimized middleware and frameworks relevant for checkout use cases from Intel.
  • Use the core checkout services to build the automated self-checkout reference modules.
  • Identify the required hardware for the intended workload.

Features and Benefits

With this reference implementation, the self-checkout stations can:

  • Recognize the non-barcoded items faster.
  • Recognize the product SKU and items placed in transparent bags.
  • Reduce the steps involved in identifying products when there is no exact match (top five choices)

How it Works  

The video stream is cropped and resized to enable the inference engine to run the associated models. The object detection and product classification features identify the SKUs during checkout. The bar code detection, text detection, and recognition features further verify and increase the accuracy of the detected SKUs. The inference details are then aggregated and pushed to the enterprise service bus or MQTT to process the combined results further.

Resources

The reference implementation includes:

  • Source code
    • Microservices
    • Benchmark scripts
    • Pre-trained models
  • Documentation
  • Learning videos (will be available in the forthcoming releases)
  • Hardware recommendation
  • Tools and libraries
  • Operating system support
  • Support for Intel architecture-based platforms

Get Started on GitHub

Performance Results

Find the latest performance results by choice of Intel® processors for the vision-enabled workloads.

View Performance Results

Report Issue or Submit Feedback

You can open an issue on GitHub to report a problem related to the reference implementation or give feedback.