Warehouse signboard identification method based on image convolutional neural network technology

Abstract

The invention discloses a warehouse signboard identification method based on an image convolutional neural network technology. The method mainly includes the following steps: first extracting a part of interest of a picture from a picture stream, later performing zooming processing on the picture, and converting into a picture of the same size; detecting a signboard picture based on a light stream method, and when a change of the picture is detected, extracting the picture at a moment when the change occurs and pictures of a previous frame and a next frame; and using the three pictures as input, utilizing a convolutional neural network to perform picture calculation, and finally judging how the signboard changes. According to the method, the area of interest in the signboard scene picture is extracted, the picture is scaled to a fixed size, the signboard change is detected, a convolutional neural algorithm is utilized to perform partitioning processing, the problem of difficult detection caused by the fact that the number of signboards is large and the circumstances are complicated in large-scale warehousing is avoided, and the test precision is greatly improved. The warehouse signboard identification method based on an image convolutional neural network technology can be widely applied to the field of warehouse management.

Description

technical field [0001] The invention relates to warehouse management, in particular to the field of warehouse sign recognition, in particular to a warehouse sign recognition method based on image convolutional neural network technology. Background technique [0002] With the continuous improvement of our country's market economy system and the rapid development of productivity, as the core of enterprise management: warehouse management has become a key factor in the competitiveness of enterprises. However, in management, large warehouses often send wrong materials and misplaced materials. It is inevitable that there are subjective elements of carelessness of the custodians, and the most important reason should be that the location and identification of the goods are not clear, and this kind of location The asymmetry with the logo, on the one hand, is due to a mistake when placing the logo, and on the other hand, it is more due to the misplacement of the logo during use. How...

AI Technical Summary

Problems solved by technology

[0004] Difficulties in storage signage management: Generally, large-scale warehouses have many types and large quantities of storage signage, and the number of warehouse managers is limited, so they cannot always observe the changes of the location signage. irreversible consequences

Images