Localize And Decode Bar Code Using Image Processing Method

In nowadays, along with the rapid development of computer technology, bar codes are used widely all around the world, due to their wonderful advantages, such as easy, convenient, low cost, high speed of saving information and transportation, et al. And bar code is the passport for a commodity entering the International market.The research of this paper is based on a entirely new shopping idea, that is, the customers donnot need to line up for the cashiers, what they just do is putting their chosen commodities into a designed self-shopping system to realize paying the bills totally by themselves. At present, the bar code scanners widely used in supermarkets are made from the principle of opto-electrical recognization method. Although this technology is mature and easy, the bar code area must be localized by manual to scan. And it requires very good quality of the printed bar codes. If there exist fragmentation, dirty marks, or something else, then the bar code scanner cannot correctly read out the information carried by the bar code. Furthermore, one dimensional laser barcode scanner cannot read two dimensional barcodes, which are used more and more widely. Using digital image processing method, the quality of the image can be improved first, to reduce or eliminate the bad effects on barcode reading caused by noise, nonuniform illumination, image blur, et al. Then decode the localized barcode area. Using image processing method, the new image barcode reader can not only read one dimensional barcodes, but also two dimensional barcodes. Recognizing barcodes via image processing method is the inevitable developing direction in the fileds of barcodes.This paper puts forward the concept of the designed self-shopping system, the encoding rules and technical characteristics of EAN-13 bar code for commodity, the basic composition parts of the bar code image reader system, et al. Analysis the problems occurring during the process of capturing barcode image. Bring forward corresponding digital image processing algorithms for image noise, nonuniform illumination, image blur, respectively, and the recognization method for bar code image.The algorthms in this paper are all simulated in MATLAB environment. Various test examples show that the algorithms for barcode localization and recognition are effective and fast. The effect of barcodes identification via image processing method is good. Compared to traditional opto-electrical method, it can solve the problems of poor quality for bar code, and the system is well-adapted.