A localization algorithm for distorted or rotated QR code

Due to the efficient encoding of Chinese characters and quickly recognition, the Matrix QR code is widely used in numerous industries. Exact localization is the premise and crux of QR code recognition while existing localization algorithms have weak robustness when the QR code is distorted or rotated. A new adaptive and robust localization algorithm for QR code is proposed in this paper. First, we use the adaptive threshold algorithm to solve the influence of uneven illumination. On this basis, we propose a localization method which is based on contour tracking and contour sifting. The experiments of several sample sets show a high detection rate while the QR code is rotated freely or distorted slightly.