Infrared Invisible QR Code Recognition Based On Image Processing And Its Application

Invisible QR (Quick Response) code is a special two-dimensional code that uses special materials to print and requires dedicated reader device to read.It can be applied to multiple areas such as logistics management, product security and authentication.Meanwhile, it is printed by the ink material that absorbs light at infrared, as a result, its imprint is colorless and does not affect the original appearance. Because the human eye can not directly observe in the bar code, it will not only be counterfeited, but also can not be copied or reproduced and can achieve a good logo and security features. However, the infrared QR Code image has problems of blurred images, uneven illumination, large noise and it can not be identified by using the ordinary QR code image processing method. In this paper, for the purpose of the QR code recognition and based on image processing of infrared invisible QR code recognition process, infrared stealth QRcode image recognition algorithm such as image grayscale, binary,QR code symbol positioning, correction and normalization is studied. This paper includes the following aspects.The first step is infrared invisible QR code recognition process based on image processing. By analyzing the characteristics of infrared invisible QR image and difficulties of infrared invisible QR code,a recognition process composed of the image acquisition, image processing and decoding steps is proposed including image processing, including gray, binarization, location and tilt correction and normalization.The second step is image preprocessing, including image gray and binarization. To sovle noise problem of infrared invisible QR grayscale image, this paper utilizes the single channels algorithms instead of gray image and the time efficiency have improved greatly.For binarization,by analyzing the infrared invisible QR code image information characteristics in the R, G, B channels, based on the Gaussian adaptive threshold algorithm, binary search strategy is propsed to solve the local optimal window value and the correction constant parameters. Experimental results show that the proposed algorithm reduces image noise, ensures the integrity of the symbol information and satisfies the subsequent processing requirements.The third step is QR symbol positioning and correction processing.To solve the failure of detection graphics-based positioning method, this paper analys the symbol of the binary QR image characteristics and proposes a positioning algorithm based on contour tracking algorithms and polygon approximation algorithm.Experiments show that the algorithm improves the positioning accuracy largly and the inclination angle of the QR code is easily calculated based on the positioning results and the version number is calculated based on the QR code characteristics on a higher speed.The forth step is image normalization and QR decoding. For the problem that the module can not be accuratly divided in average normalization algorithm which is leading to uncorrectly decod QR code, this paper presents an adaptive grid normalization algorithm to achieve the standardization of the QR Code. After the implementation of decoding algorithm using preprocessed QR code image, the recognition rate has rearched to90.3%and basically satified the preliminary practical requirement.The last one is application of security and identification information system. Combined with the results of infrared QR image recognition and analysis the design philosophy of security and identification information system, system architecture and main functions is designed and programmed.