Study On Image Detection For Two-dimensional Barcode QR Code

With the rapid development of the world economy, bar code technology is used more and more widely in all trades and professions. The bar code technology develops on the basis of computer industry, photoelectric sensor technology and advanced communication technology. The diverse information of modem commodity requires the bar code to denote more information in limited geometry space in order to meet every kind of market demand. The appearance of two-dimensional barcode surmounts many limitations of one-dimension barcode. Nowadays, the technology of two-dimensional barcode has been applied extensively overseas while it is just used not long ago in our country. And there are differences in some crucial skills between the domestic and foreign two-dimensional barcode. Because QR code is able to express Chinese character information, it is fit for being utilized in our country. Considering many problems of QR code in actual application, such as the stability, accuracy, reliability, and so on, we should research on it so as to improve further the performance of two-dimensional barcode in practical application.This paper firstly introduces the process of collecting QR image which is completed with the assistance of camera, light source, PCS, main control board, signal sensors, encoder and so on. Next, it focuses on the preprocessing of QR code image, including the binary image processing, image positioning, filtering and so on. And it analyzes the encoding and error correction principle of QR code in details so as to design a series of image recognition software of QR code. In order to accelerate to identify the image, the method of run-length coding is adopted to handle the image of QR code by means of the statistical features of image so as to realize the image segmentation conveniently. Each line of the binary image of QR code is encoded by run-length coding and a series of images consisting of data stream can be formed. The related parts among these lines come into being an area. As a result, every part of the whole image of the QR code can be divided. The areas of image include data area, coding region and so on. These procedures make good preparation for handling the following image processing. Afterwards, the horizontal and vertical projections of the QR code image are generated. The projection point in which pixels information suddenly increases is taken as the center and the average value of the variance of the rest scattered pixel to this projection point is taken as the threshold value, the marginal location of the QR code can be determined and the image edges are able to obtained after filtering certain noise. Better graphics and fine images can be achieved using the above preprocessing algorithm.The error correction decoding of QR code data is the critical step in the process of decoding. This paper elaborates on the decoding design on the basis of coding and decoding principle of QR code. After the image information being transformed into identifiable digital forms, we can make sure of identifying the digital information correctly through correcting digitals. Moreover, we investigate the decoding algorithm of the QR code image. The computer program in this topic has debugged successfully in PC and it possesses good portability. It can also be used to test the decoding system through the realized coding system. This decoding system can decode efficiently the QR code images with accuracy. It has excellent performance in use.