An aesthetic QR code solution based on error correction mechanism

• Published in: The Journal of systems and software Vol. 116; pp. 85 - 94
• Main Authors: Li, Li, Qiu, Jinxia, Lu, Jianfeng, Chang, Chin-Chen
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.06.2016; Elsevier Sequoia S.A
• Subjects: 2D barcode; Aesthetic QR code; Algorithms; Bar codes; Coding theory; Computer programs; Decoding; Error correction; Error correction & detection; Image detection; Mathematical analysis; Modules; Saliency detection technology; Software; Studies; Saliency detection technology; 2D barcode; Aesthetic QR code
• ISSN: 0164-1212; 1873-1228
• DOI: 10.1016/j.jss.2015.07.009

•Our aesthetic QR code generation algorithm is based on error correction mechanism, use the characteristics of QR code.•In order to highlight the important regions of background image, we combine with the saliency technology to detect the significant regions.•Compared with the existing methods, our algorithm can maximize the changeable areas.•This algorithm has better aesthetic effects, while keeping the rate of successful decoding. QR code(Quick Response Code) is a popular two-dimensional matrix that randomly consists of black and white square modules. While the appearance of QR codes are often visually unpleasant, it leads to the increasing demand for the aesthetic QR codes. However, it may turn out to be unreadable if changes to the modules of the QR code are inadequate. Therefore, to resolve this conflict, we propose a method to generate an aesthetic QR code, which is based on the RS(Reed–Solomon) error correction mechanism in QR code encoding rules. First, according to the characteristics of the QR code, we mark the positions of codewords as codeword layout. Then, we detect salient regions of the background image to generate the saliency map. The next step is to combine it with the saliency map and codeword layout to calculate saliency values, then sort and select proper codewords as changeable regions. Finally, we propose the hierarchical module replacement rules. The theoretical maximum value of the changeable areas is the redundancy capacity T of RS error correction. Compared with the existing methods, our algorithm can maximize the changeable areas and highlight the important regions of background image. This algorithm has better aesthetic effects, while maintaining the rate of successful decoding.