Robust and fast QR code images deblurring via local maximum and minimum intensity prior

QR codes, as information carriers, have been extensively applied in numerous fields. However, relative movement between imaging targets and the camera can lead to severe motion blur, which hinders the recognition of the QR code. Achieving high-quality deblurring while maintaining efficiency is a challenging task in practice. In this paper, we propose a robust and fast deblurring method for QR code images using the local maximum and minimum intensity (LMMI) prior. First, in the process of image blur degradation, a general feature emerges: local maximum intensity tends to decrease, while local minimum intensity increases. This characteristic can be effectively harnessed in the deblurring process of QR code images. Second, the novel algorithm flexibly binarizes the LMMI within the maximum a posteriori (MAP) framework, as opposed to directly applying the semi-quadratic splitting algorithm, resulting in a significant enhancement in computational efficiency. Finally, latent image estimation and kernel estimation are alternately solved within the MAP framework until convergence. Experimental results reveal that the proposed method exhibits considerable improvements in both deblurring effect and computational efficiency.