Robust and Fast Decoding of High-Capacity Color QR Codes for Mobile Applications

摘要

The use of color in QR codes brings extra data capacity, but also inflictstremendous challenges on the decoding process due to chromatic distortion,cross-channel color interference and illumination variation. Particularly, wefurther discover a new type of chromatic distortion in high-density color QRcodes, cross-module color interference, caused by the high density which alsomakes the geometric distortion correction more challenging. To address theseproblems, we propose two approaches, namely, LSVM-CMI and QDA-CMI, whichjointly model these different types of chromatic distortion. Extended from SVMand QDA, respectively, both LSVM-CMI and QDA-CMI optimize over a particularobjective function to learn a color classifier. Furthermore, a robust geometrictransformation method is proposed to accurately correct the geometricdistortion for high-density color QR codes. We put forth and implement aframework for high-capacity color QR codes equipped with our methods, calledHiQ. To evaluate the performance of HiQ, we collect a challenging large-scalecolor QR code dataset, CUHK-CQRC, which consists of 5390 high-density color QRcode samples. The comparison with the baseline method [2] on CUHK-CQRC showsthat HiQ at least outperforms [2] by 188% in decoding success rate and 60% inbit error rate. Our implementation of HiQ in iOS and Android also demonstratesthe effectiveness of our framework in real-world applications.