Open AccessAn Empirical Study on Smart Steel Stockyard Management System through Character Recognition on Steel Materials using CNN
Author(s) -
Gi Yeong Cho,
Nam-Hyun Yoo
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1117/1/012025
Subject(s) - shipyard , steel mill , computer science , plan (archaeology) , engineering , shipbuilding , metallurgy , materials science , geology , paleontology , archaeology , history
In shipyards wherein large ships and plants are constructed, the timely supply of steel is a critical factor that can decrease the production costs and shorten the construction period. However, the standard size of the steel materials utilized in shipyards exceeds at least 10m, and the weight of the steel is heavy; therefore, it is difficult to readjust the steel once it is deployed. Moreover, if any part of the production plan is changed, the steel that has been deployed according to the existing plan must be relocated. However, when cumbersome, heavy steel is vertically stacked in several layers, it is difficult for a shipyard worker to directly identify the desired steel’s location. Therefore, it is not easy to spot the location of the steel to be found. In cases such as this, one attempts to recognize characters on the steel’s surface using OCR, but it is difficult to apply this to an actual system as the recognition rate is only about 50 to 70% on average. Therefore, through this study, we intend to develop a model that recognizes characters printed on the steel’s surface using CNN and to check whether it can be employed in a smart steel stockyard management system.