Bar Code Decoding in a Camera-Based Scanner: Analysis and Algorithm

Abstract

The goal of this work is to provide a foundational understanding of the resolution requirement for camera-based bar code scanning systems and to develop an effective algorithm for decoding. The main theoretical question we wish to address is how well resolved must the bar code image be for the system to be able to decode it. To facilitate the analysis we consider the UPC bar code, which is widely used in retail and commerce. We further assume that the pixels of the camera are aligned with the bar code, making the problem one-dimensional. We find that if the narrowest bar in the bar code is larger than two-thirds the size of the camera pixel, it is possible to uniquely determine the encoded message in the bar code. The result, which exploits the symbology of the UPC bar code, shows that under-resolved bar code images can be decoded. To further gain insight into the robustness of the decoding process, we perform numerical experiments.