The Research On The Technology Of Invisible QR Code Recognition

Because of its advantage of large information capacity, high reliability, fastrecognition speed, wide recognition range, various encoding categories, quickresponse code (QR code) has been widely used in many fields. Invisible QR codemakes QR code inherits the brilliant security capabilities from invisible printing.Invisible QR code images are more difficult to separate the QR code and thebackground. This paper puts emphasis on how to separate the QR code informationand the background from the invisible QR code, how to read the data from the QRcode and improve the accuracy of invisible QR code recognition. Throughtheoretical analysis and experiments, this paper mainly studys the invisible QR coderecognition technology, the main content of this paper is as follows:1It is difficult to separate image background from the QR code module for thecolor distortion and the uneven illumination when shooting the invisible QR. Tosolve this problem, this paper firstly calculates the grayscale value of the QR codeaccording to the situation of the QR code information in invisible QR code image’sR channel is outstanding. After that, using the adaptive grayscale image binarizationmethod, this paper separates the QR code and the background and recognizes thedifferent depth of modules.2Because the background of image is complex and there are some errors whenrecognizing the different depth of modules, it is hard to locate the QR code image.According to characteristics of the invisible QR code image, this paper modifies theimage location method in QR code rules to improve the accuracy of finding QRcode location detecting images.3The sampling images can be distorted when collecting the QR code images.With reference to the image sampling method in QR code rules, this paper proposesa method to solve the sampling image distortion. In this method, the first step is tofind the correction graphics center, and the next step is to sample the QR codeimages using the area network established by the location of the correctiongraphics.4To improve the accuracy of recognizing the invisible QR code images, thispaper chooses a error correction code word with correction level H to correct thedata. This paper also implements the RS error correction code mapping table andinverse mapping table in the primitive elements of GF(28).5To further enhance the security capabilities of the invisible QR code, thispaper chooses the AES method to encrypt the input data.