Research Of Location And Correction Algorithm Of QR Code Based On Image Processing

As the most widely used matrix two-dimensional bar code, QR (Quick Response) code not only has the advantages of high capacity, strong correcting ability, but also can display Chinese characters and response to scanning recognition at arbitrary angle, which has been widely applied in logistics management, product marketing, electronic ticketing etc.. In the application of the two-dimensional code technology, artificial positioning such as hand-held gun sweep code scanning, mobile phone focus scanning etc. are still widely used, but these methods are not applicable to the heavy workload and single content industrial fields like intelligent express parcel sorting links. The continuous development of information technology provides an intelligent solution for the industrial production of this kind of industry which needs a lot of manpower. One dimensional bar code identification technology is still widely used in the existing express parcel sorting system. With the maturity of the two-dimensional code technology, two-dimensional code technology has a better prospect and a wide range of applications in the intelligent sorting system.Aiming at the existing difficulty of positioning and identification in the application of two-dimensional code, some of the current two-dimensional code positioning and recognition algorithms were studied, and as well the advantages and disadvantages of these algorithms, then a positioning and correction algorithm for QR code using symbols features was proposed. The algorithm locates all the high contrast areas with partial gray maximum difference sampling method, and then screens the two-dimensional code region and realizes localization of QR code region according to the texture feature and the symbolic features of the two-dimensional code, and segments the symbol to reduce the amount of computation of the follow-up operation. According to the characteristics that the center points of three position detection graphics of standard QR code symbol constitute an isosceles right triangle, the tilt angle of the QR code was confirmed by judging the shape distortion of the triangular, and then spinning processing can be done, finally, the sampling network correction of traditional decoding methods was improved to enhanced the decoding rate.Verification of the algorithm was carried out with MATLAB, verification results show that, the algorithm can quickly and accurately positioning and recognition of QR code at any angle, and just required low computational complexity, which is more convenient and accurate compared to the traditional edge detection and Hough transform. In addition, hybrid programming of the algorithm was done base on Microsoft Visual studio2010 and MATLAB2010, and an intelligent sorting system for express parcel was modeled base on two dimensional code real-time recognition technology, and experiments were carried out on the sorting site for the algorithm. Experimental results show that, the correct decoding rate can reach to over 95.74%, which can meet the needs of industrial field such as express parcels intelligent sorting links.