Loss Prevention Retail Reference Implementation

Overview

The Loss Prevention Reference Implementation provides critical components to build and deploy solutions to prevent store theft. This reference implementation focuses on theft at shopping aisles and checkout stations. This reference implementation provides pre-configured loss prevention pipelines optimized for Intel® hardware.

Requirements 

To build the Intel® Loss Prevention 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.
• Utilize services developed for the self-checkout reference implementation.
• Use the core checkout services to build the loss prevention reference modules.
• Identify the required hardware for the intended workloads.

Features and Benefits

With this reference implementation, the retail stores can:

• Monitor persons of interest
• Track products that customers might hide (Recognize suspicious customer activity).
• Identify price switching.
• Keep track of the items skipped in the basket and shopping cart.

How it Works

This reference implementation identifies video streams projected by various types of cameras, such as the overhead cameras at the isles, shelf monitoring cameras, and cameras in the self-checkout area. Different AI models with varying resolutions and precisions drive this reference implementation.