Commodity identification method and device based on neural network, and self-service cashier desk

Abstract

The invention discloses a commodity identification method and device based on a neural network, and a self-service cashier desk. The method comprises: obtaining an image containing a commodity to be detected; and inputting the image containing the commodity to be detected into an identification system based on a neural network; and enabling the identification system based on a neural network to output the information of the commodity to be detected. The method obtains a product image through an ordinary camera, obtains the product information by using an image recognition algorithm based on the neural network without the need of the third-party identification, a user only needs to put the purchased product under the camera to obtain the image, and then the identification is realized. The method and device are low in use cost and high in identification accuracy.

Description

technical field [0001] The invention relates to a neural network-based product identification method, device, and self-service cash register, and belongs to the field of deep learning neural network and image recognition. Background technique [0002] There are two main types of commodity machine identification in the existing self-service settlement scenarios: [0003] The first is based on RFID electronic tags (Radio Frequency Identification, also known as radio frequency identification) to identify settlement methods. First, specify the corresponding relationship between commodities for each electronic tag with a unique ID in the database, and then post the electronic tag to all On this type of product on sale. During settlement, the unique ID of the electronic tag is read out through the card reader, and the information of the product is queried in the database based on the ID, so as to complete the "identification" of the product and perform settlement. This settlement ...

AI Technical Summary

Problems solved by technology

This method requires the user to scan in person, which is prone to various scanning misoperations and increases the cost of use

For example, multiple scans of a product lead to repeated counting, and it is difficult to scan correctly after the product barcode is deformed, etc.

Simultaneously, this method also has the problem of poor anti-theft effect

Images