Visual inventory inspection using optical character recognition

This paper presents a research on developing a visual inspection of character base image for automatic inventory control system by using an optical image processing method. Usually, bar code method has been used as an input device for inventory control system. However, data or information that included in a bar code may not sufficient for the purpose of inventory controlling. Therefore, for the purpose of efficiency, character recognition for label stock in inventory controlling system will be an advantage in providing extra information regarding stock in industrial computing. This system derives from 2 major parts. The first part is used to extract the area of stock label in a grayscale picture, and then try to recognize the text inside a label. Whereas the second part, it record the in-out activities of an inventory using the information from the first part. Research for the recognition system is concentrate on the capital letters and the numbers only.