Cashiering method for identifying commodities based on neural network and self-service cashier desk

Abstract

The invention discloses a cashiering method for identifying commodities based on a neural network. The method comprises the following steps of acquiring commodity images of commodities to be detectedwhich are placed in different grid grooves, wherein each grid groove is used for containing commodities with the same kind; acquiring weighing weight of the commodity to be detected in at least one grid groove; inputting the commodity image into an identification system based on the neural network, wherein the identification system based on the neural network outputs commodity information in the commodity image; and outputting an identification result according to the weighing weight and the commodity information. According to the method, the commodities with the same category are identified by a neural network system, so that the identification accuracy can be improved; the weighing weight of the commodity to be detected can be applied to an identification process, so that the identification process can be simplified, and the identification efficiency can be improved; and furthermore, the weight of the commodity to be detected can also be verified, so that the identification accuracyis further improved.

Description

technical field [0001] The invention relates to a cash register method for identifying commodities based on a neural network and a self-service cash register, belonging to the field of image recognition. Background technique [0002] In order to improve the convenience of life in the community, unmanned convenience stores have been built around many communities, and self-checkout machines are installed in unmanned convenience stores. [0003] Most of the existing self-service checkout machines automatically scan codes for customers to check out. Customers need to scan codes for each item, so the checkout efficiency is very slow, especially when there are many customers in the convenience store, which reduces the customer's experience in unmanned convenience stores. feel. [0004] Now there is also a self-checkout machine based on neural network recognition technology. When using the checkout machine, the customer places all the items on the cash register, and the automatic ...

AI Technical Summary

Problems solved by technology

Then, because all the items are piled up on the cashier counter, it is easy to pile up or overlap, so the camera cannot fully recognize all the products, so it is necessary to repeatedly strengthen the training of the neural network, and repeated reinforcement training of the neural network takes a lot of time. time, leading to a significant increase in the time cost of self-checkout machines on the market

And the existing self-checkout machine based on neural network recognition technology can only check out the goods sold by counting, so the goods that are checked out by weight cannot be sold in unmanned convenience stores, which reduces the number of goods that can be sold in unmanned convenience stores. Variety of goods, reducing convenience

Images