Commodity automatic identification method, system and device and storage medium

Abstract

The invention discloses a commodity automatic identification method, system, device and a storage medium. The method comprises the steps of obtaining a first commodity image; inputting the first commodity image into a trained deep learning model to carry out feature map extraction processing; obtaining a feature map of the first commodity image, and obtaining commodity position annotation data ofthe first commodity image from the feature map, then using the feature map by the deep learning model to carry out commodity type identification processing on an image block corresponding to the commodity position annotation data of the first commodity image, and outputting the identified commodity type. Therefore, the embodiment of the invention has the advantages of high intelligent automation degree, low purchase cost of input equipment, high processing efficiency and accurate identification, and can be widely applied to the field of automatic retail.

Description

technical field [0001] The present invention relates to data identification processing technology, in particular to a commodity automatic identification method, system, device and storage medium. Background technique [0002] Explanation of technical terms: [0003] RFID: Radio Frequency Identification, radio frequency identification technology, also known as radio frequency identification, is a communication technology that can identify specific targets through radio signals and read and write related data without the need to establish mechanical or optical contact between the identification system and specific targets . [0004] With the development of science and technology, more and more scenes have gradually realized automation or semi-automation. In the context of new retail, automated retail has also received extensive attention in recent years. The concept of new retail is based on the reconstruction of the three elements of business - people, goods, and places. It...

AI Technical Summary

Problems solved by technology

[0005] At present, the commodity identification link of new retail mainly uses barcode identification and RFID radio frequency identification technology, among which: 1. Using barcode identification technology, which mainly adds barcode printing marks to each commodity package, which greatly facilitates the circulation of commodities. Now every Almost all commodities have their corresponding barcodes. However, due to the uncertainty and randomness of the barcode printing position, it is often necessary to manually find the barcode to assist the machine in identifying the barcode during settlement, and the degree of automation is low; 2. Using RFID Radio frequency identification technology, which is similar to barcode, needs to affix an RFID label to each product, each label has its specific number corresponding to a specific product one by one, and the product can be identified through wireless signal communication, although its It solves the problem of contact identification, but due to the interference between wireless signals, the identification of multiple commodities at the same time still has a high error rate, and it needs to affix an RFID tag to each commodity, and the cost of RFID tags is relatively high. High input cost of using RFID technology, and it has disadvantages such as difficulty in recycling, so it is difficult to meet the reciprocal conditions of commodity profits, so it cannot be widely used

Images