Performance Evaluation of Corrupted QR Code Scanners

Quick Response (QR) Code is touted for its ease and convenience. It has therefore seen tremendous applications today. The application of the codes for education, healthcare and marketing cannot be overemphasized. However, like other symbology technologies, the use of QR Codes may pose a challenge for users when scanners cannot scan codes at crucial moments. Unsuccessful or invalid scanning may be due to poor QR Code, the camera quality of the smartphone, corrupted QR Code image and poor implementation of scanning algorithm. In this paper, an evaluation of 10 highly rated QR code scanners is performed. Specifically, scanners resilience to corrupted images is evaluated. NeoReader QR outperformed the other scanners, while Kaspersky QR Scanner performed least. Generally, it was observed that scanners perform well when the number of characters encoded in the QR code is few, whether codes are corrupted or not. QR Codes with a high number of characters encoded in them are sensitive to noise. Again, scanners can handle Gaussian noise better than salt-and-pepper. Integration of median filter in the scanning pipeline to help suppress salt-and-pepper noise can make QR Code scanners robust.