Commodity shelf vacancy identification early warning method and system based on visual algorithm

Abstract

The invention provides a commodity shelf vacancy identification early warning method and system based on a visual algorithm. The identification early warning method comprises the following steps: extracting and selecting a plurality of training images; performing manual labeling on each training image to obtain corresponding labeling vacant area data, labeling commodity label frame data and labeling commodity identification character data; a vacant area detection model, a commodity label box detection model and a label character recognition model are constructed and trained; obtaining current commodity label box data, current commodity identification characters and current vacant area data corresponding to the current image through the trained commodity label box detection model, label character identification model and vacant area detection model, and calculating to obtain a current vacant area proportion in an area in which the corresponding commodity is stored; judging whether the current vacant area proportion is greater than a preset vacant area proportion or not; and when the judgment result is yes, sending early warning information containing the proportion of the current vacant area and the corresponding commodity identification characters.

Description

technical field [0001] The invention belongs to the technical field of intelligent commodity management, and in particular relates to a visual algorithm-based method and system for early warning of commodity shelf vacancy identification. Background technique [0002] In daily life, some shopping places such as stores and supermarkets, due to the large number of shoppers and the scattered shopping, can accurately and timely find the vacant goods and remind the staff to replenish the goods in time, which can not only better meet the needs of customers, but also It can boost the profit of the supermarket. [0003] Usually, shopping places employ store clerks to frequently count the shelves, organize and report them, and then replenish the goods. This may cause problems such as inefficiency caused by the delay of the clerk in the inventory, and frequent inspections by employees beside the shelves. SUMMARY OF THE INVENTION [0004] The present invention is made to solve the ab...

AI Technical Summary

Problems solved by technology

[0003] Usually, shopping places hire clerks to check the shelves frequently, arrange and report them, and replenish goods. This may cause problems that the clerks do not count in time, which leads to inefficiency, and requires frequent inspections by employees next to the shelves.

Images