Design of a barcode identification system

Modern barcode readers need to decode many symbologies. It is inefficient for the readers to read and decode a barcode symbol in the try-and-error manner. That is, it is inefficient to decode a given symbol by testing all symbologies supported by the reader. A more efficient decoding scheme is to identify the symbology used by the target symbol. Then the decoder in the reader uses the correct symbology for decoding. In this paper, we propose a method of identifying the QR and Aztec barcodes. We use the connected component labeling algorithm to compute the tag for each connected component, then search for the innermost connected component of the finder patterns in the tag image. Experimental results show that the method can identify rotated or defocused barcode images.