Deblurring QR Code Images using Local Maximum and Minimum Intensity Prior
Rushi Jin, Yuanhaoji Sun, Bin Xu, Dazhi Zhang, Yong Guo, Yongsheng Li, Jinhua Li, Bo Zhang, Kai Liu · 2023
We propose a robust and fast QR code deblurring method using local maximum and minimum intensity (LMMI). The motivation for this work stems from a common characteristic observed during the image blur degradation process: the local maximum intensity tends to decrease, while the local minimum intensity tends to increase. This characteristic can be effectively utilized in the deblurring process of QR code images. Our algorithm binarized LMMI through a simple thresholding step, which improved the computational efficiency compared with the semi-quadratic splitting algorithm. The latent image estimation and kernel estimation are alternately solved within the maximum a posteriori framework until convergence. Experimental results show that our method significantly improves both deblurring effect and computational efficiency.