A low-complexity pre-processing system for restoring low-quality QR code images

In this paper we propose a complete preprocessing system for restoring low-quality Quick Response code images. The target application for the system is to restore and decode a code image in a cell phone when the image is shown in another cell phone display. Therefore, the system design focuses on low complexity solutions that are effective for a variety of capturing and displaying devices. The system has been designed to deal with low contrast, blurred, rotated, and deformed code images, while keeping complexity low enough to run in real-time. The proposed system takes advantage of intrinsic Quick Response code characteristics in order to restore the code properly and efficiently. The system has been assessed on a realistic database composed of 533 images captured by a cell a phone from several different cell phone displays. After the proposed preprocessing, 97.4% of the images in the database were correctly decoded.