Research On The Recognition Technology Of Quick Response Code

Quick Response code is a sort of two-dimension code, also called QR code. In recent years, it was applied in our daily life more and more extensively, especially with the development of mobile internet, QR code makes our life more convenient everywhere. QR code was been invented in 1994 by the Japanese company Denso wave which is a subsidiary of Toyota. QR code has a lot of advantages, like big storage capacity, high security and stability, rich data-type, etc. So it was applied widely in many industries, like logistics, warehousing, library, postal, transportation, etc. After the development of the mobile internet, QR code gained a significant presence in the mobile internet, now we can see many mobile applications with barcode scanning function since QR code make the functions of mobile application easier like inquiry, pay, download, etc. This tiny symbol deeply affects our life gradually.Of cause with the development and promotion of QR code there is also have a series of relative problems. In the practical application, QR code situates complex environment: digital screens, print, outdoor ads, etc. Certainly it will affect the acquisition of QR code image, there may be some problems occur like tilt distortion, geometric distortion, uneven illumination; it may lead the QR code cannot identify accurately.This thesis studied around these issues may occur in the acquisition of QR code. Based on the uneven illumination, tilt distortion, geometric distortion, to correct these problems occur in the QR code image, a series of correction methods, namely, graying, filtering, binarization, tilt correction, geometric correction and sampling, study the existed algorithm of each pretreatment process. The geometric distortion and sampling are mainly studied in this thesis, based on the tilt correction the key point in the geometric distortion is the location of four vertices of QR code. A new method of vertex location is proposed in this thesis, it is simpler and easy to operate compared with the traditional methods of edge detection and Hough transform. After the geometric correction, some QR code still with a problem that the size of submodule is nonuniform which is distorted seriously; the traditional method cannot sampling accurately, so based on that situation, this thesis proposed a sampling method based on adaptive match, this method can sampling these QR code with nonuniform submodule accurately and improve the efficiency, it was worth researching in the practical application.Finally summarized the whole thesis and made prospect for the future research.