Authors:
Patricia Hogan, Daniel Goins, Jessica Snead, Susan Brosnan
Abstract:
Some companies offer CT scanning poles that event venues use to scan guests to look for firearms. This solution proposes using CT scanning poles and associated AI that is trained to recognize items so if a shopper has added items from the store to the shopper's person, pockets or bags they could be identified. This could be used to aid loss prevention in stores.
Background:
It is hard to detect if a shopper has hidden unpaid items on their person, in their pockets, or in their bag. Our idea is to use the CT scanning poles that are used at some event venues and train the AI associated with the CT scanning poles to detect if the shopper has hidden items on their person without paying for the items.
Description:
At the Durham Performing Arts Center in Durham North Carolina, and in Walnut Creek Amphitheater, the people who run those entertainment venues have moved to using CT poles that guests walk past as they enter the venue. The CT poles scan the guest and their bags as the guest walks past the poles without the guest needing to disrobe or needing to open their bag for a security person to search through. The CT poles take images of the guest and their bags as the guest walks by. The images are fed to an AI image recognition system that has been trained to recognize firearms or other weapons and to circle them in the image and send an alert to security personnel with a picture of the person and where the weapon is.
Our idea is to use this same technology but train the AI model differently.
On the way into store, the shopper and their bags are scanned as they walk past the CT poles. The images are saved for the short term. As the shopper approaches the exit of the store, or the checkout area, they walk past another set of CT poles. These CT poles scan the shopper and their bags again. The second set of images is compared to the first set of images to determine if the shopper has hidden any items that they have not paid for on their person. This could be used in a regular store, but it would also make a great addition to a frictionless store where video recognition is trying to recognize all the items the shopper picks up. The CT scan could be used to confirm the shopper has the items the system has charged them for and could detect any missing items that were paid for, but the shopper does not have with them or detect extra items the shopper has not paid for.
Enabling Technology:
https://www.evolvtechnology.com/industries/ticketed-venues
https://xtractone.com/walk-through-metal-detectors-vs-xtract-ones-smartgateway/?utm_term=evolv%20technology&utm_campaign=SEP-Search
Supporting Art:
https://milliscan.fi/2023/04/25/milliscan-in-loss-prevention-conference
TGCS Reference 00114
