•  
  •  
 

Abstract

Multi-symbology barcode location is a practical and challenging issue. Image-based barcode recognition techniques are robust and extendable approaches for reading versatile one-dimensional (1D) and two-dimensional (2D) barcodes. Most methods may work for a single 1D or 2D barcode or rely on a unique pattern to locate a single 1D or 2D barcode. This work proposes a general localization framework for extraction of real barcodes under complex backgrounds when multiple symbology types exist in the same snapshot for 1D barcodes, 2D barcodes, or both. There are five steps of the proposed algorithm: image grayscale conversion, adaptive thresholding, application of a modified run length smearing algorithm, connected-component labeling, and barcode verification. Experimental results indicate that the proposed approach can locate multiple barcodes and multi-symbology barcodes in complex backgrounds with acceptable accuracy.

Download
COinS