QR code image detection using run-length coding

Abstract

In order to improve the practical application property of the two-dimensional barcode Quick Response (QR) code, we investigate the coding and decoding process of the QR code image. Run-length coding is applied to binary QR code image so as to accelerate the identification of QR code image. The QR code is transformed into many runs of data in alternate pixels of black and white. The related runs of data among adjacent rows are formed a unit module. After the whole image has been scanned, all of such modules in binary QR code image can be generated accordingly. With a noisy QR image captured by an industrial camera as an example, the experiments of image binarization, image seeking and localization adjustment are accomplished in sequence. Also the error correction algorithm is discussed in detail. A decoding system of QR code is designed and the online detection experiments are carried out. The satisfied results are achieved.