Automatic vending recognition method based on dynamic vision

Abstract

The invention relates to the technical field of automatic vending, in particular to an automatic vending recognition method based on dynamic vision, which comprises the following steps: A, shooting a shopping process and generating shopping video data; B, continuous multi-frame images in the shopping video data are calculated through a target extraction algorithm; c, identifying and extracting a target area and segmenting a non-target area; d, carrying out fixed-number key frame sampling on each video frame in the shopping video data; e, inputting the sampled key frames into a commodity identification model for comparison processing; f, identifying the shopping behaviors by combining a behavior algorithm, and obtaining a final shopping result. According to the automatic vending machine, automatic vending recognition can be rapidly and accurately achieved, the requirement for the placement position of the commodity is not limited, placement of the commodity is facilitated, the overall space utilization rate of the vending machine can be increased, and the automatic vending machine can be well suitable for unmanned automatic vending machines.

Description

Technical field: [0001] The invention relates to the technical field of automatic vending, in particular to a dynamic vision-based automatic vending recognition method. Background technique: [0002] With the development of technology, the retail industry has flourished in recent years, and some unmanned vending cabinets have appeared on the market, which can realize unmanned automatic vending. [0003] Most of the existing unmanned vending cabinets realize automatic vending recognition based on static recognition. For example, the Chinese invention patent with the application number CN201910708702.8 in the prior art discloses a vending machine based on an image acquisition device. It obtains the original picture through the camera when the user opens the cabinet door, and obtains the current picture through the camera when the user closes the cabinet door. By comparing the original picture with the current picture, the types of commodities sold and the quantity of commoditi...

AI Technical Summary

Problems solved by technology

[0003] Most of the existing unmanned vending cabinets realize automatic vending recognition based on static recognition. For example, the Chinese invention patent with the application number CN201910708702.8 in the prior art discloses a vending machine based on an image acquisition device. It obtains the original picture through the camera when the user opens the cabinet door, and obtains the current picture through the camera when the user closes the cabinet door. By comparing the original picture with the current picture, the types of commodities sold and the quantity of commodities corresponding to each category of commodities sold are obtained. ; It not only needs to install 1-2 cameras on each shelf in the vending cabinet, but also in order to be able to take a full picture of the goods on the shelf, the camera installation distance has a certain height requirement for the goods placed, that is, adjacent A large space is required between the two-story shelves, resulting in low overall space utilization and fewer commodities; in addition, the requirements for the placement of commodities are high, and the commodities need to be placed neatly and cannot be stacked , not only is it more troublesome to place, but also can only place fewer products

Images