Vision-Based One-Dimensional Barcode Detection

The correct barcode detection is an important task to identify items and products from images. It is the cornerstone of the barcode decoding stage. Simply, incorrect detection leads to incorrect recognition. There are several studies which presented various level of accuracy based on the utilizing image processing or artificial neural networks. This paper proposes an image processing-based method to detect the one-dimensional barcode. The method proposed has succeeded when examined two published datasets of 215 and 1055 barcode images to reach accuracy rates of 86% and 85.8%, respectively. The yielded results of this paper have exceeded the results of the previous studies of the publishers of the datasets.