Automated Retail Checkout Scanner Demo

Demo Background

Self-checkout in retail and grocery stores have streamlined the customer experience. Simple and straightfoward kiosks let customers scan their orders and pay in a low-contact environment. This sector continues to expand into fully automated systems such that customers can scan multiple items at a time without needing to find barcodes and orient their items properly. These systems carry further benefit by speeding-up the process, reducing contact surfaces, etc.

This demo shows an automated retail checkout scanner using the AM62A processor, on which a deep learning model runs to detect 12 different types of food objects like bananas, apples, chip bags, soda cans, etc. The video below shows the demo as its running during at the Embedded World tradeshow in 2023.

The github repo hosting the source code and user guides contains everything needed to reproduce or even expand on the demo. There are instructions to run the demo as-is on the AM62A.

How to get started

Please find the following resources for reproducing the demo. This requires:

1. Get an AM62A starter kit EVM and a USB-camera or an IMX219 camera module with MIPI-CSI2
2. Download the Edge AI Linux SDK
3. Load the Edge AI Linux SDK via an SD card using the quick start guide
4. Log into the EVM through a network connection
5. Clone the git repo for this demo onto the EVM (or copy all files to the SD card)
6. Run the retail-shopping/run_demo.sh script

The run_demo.sh script will automatically download a pretrained and precompiled model needed for this demo. Additional resources are linked for diving deeper into how the application works and how Edge AI tools were used to enable this application. The github repo includes information about the objects used to train the deep learning model, a link to download the labelled dataset, guidance on training and compiling the model for the AM62A architecture (or any other AM6xA Edge AI SoC from TI), and an in-depth explanation about how the application was developed around the model using gstreamer and python3.

Additional Resources for the Retail Checkout Demo

General Edge AI and AM62A resources

Please find the following resources related to the AM62A, the retail demo, and Edge AI.