Anti-theft early warning system and method for supermarket commodities based on artificial intelligence

Abstract

The invention belongs to the technical field of commodity theft prevention, and particularly relates to an anti-theft early warning system and method for supermarket commodities based on artificial intelligence. The method comprises the following steps: when a user purchases commodities, a shopping cart logo and commodity images are obtained; and an artificial intelligence learning algorithm is applied to conduct image identification on the commodity image, conduct statistics on the number of purchased commodities; when the user settles the commodities, the shopping cart logo and bar codes ofthe settled commodities are obtained, and the quantity of the settled commodities is counted; the quantity of the purchased commodities and the quantity of the settled commodities are checked. If verification is inconsistent, the shopping cart logo is set to be an abnormal shopping cart logo; if an abnormal shopping cart is detected when the user passes through an access control terminal, an alarmis issued. The anti-theft early warning system and method for supermarket commodities based on the artificial intelligence check the quantity of purchased commodities put into the same shopping cartand the quantity of the settled commodities, and alarms the abnormal shopping cart passing through the access control terminal to achieve a purpose of theft prevention.

Description

technical field [0001] The invention belongs to the technical field of commodity anti-theft, and in particular relates to an artificial intelligence-based supermarket commodity anti-theft early warning system and method. Background technique [0002] Pasting anti-theft labels and manual identification are two common measures taken by supermarkets to prevent goods from being stolen. Sticking anti-theft labels means that merchants paste electronic labels that can be identified by radio frequency or audio-magnetic on the inside or surface of commodity packaging, and set up radio frequency or audio-magnetic erasing devices and identification devices at cash registers and supermarket exits. Erase the radio frequency or audio-magnetic mark, and check the erased radio frequency or audio-magnetic mark when the product goes out to find the product that has not been erased, so as to achieve the purpose of commodity anti-theft. [0003] Pasting electronic tags has the following disadv...

AI Technical Summary

Problems solved by technology

Manual identification is to deploy safety and loss prevention personnel at the entrance of the supermarket to check whether the products purchased by customers are consistent with the paid list. However, manual verification is intensive and inefficient, and it is easy to miss or make mistakes.

[0004] In recent years, there are also patents that try to prevent theft by combining weight and taking pictures. The basic idea is to record the body weight of customers when they enter the supermarket, and identify the products purchased by customers by taking pictures at the checkout and obtain the total weight of purchased products. Customers Weigh again when going out of the supermarket, so as to judge whether the customer has stolen behavior by comparing the weight of the customer when he enters the supermarket plus the shopping weight and the weight when he finally walks out of the supermarket. However, due to multiple factors such as changes in body weight, The operability of this method is low

Images