A new method of angle-robust multiple 1D-barcodes detection

This paper presents a new method to solve the problem of multiple 1D-barcode detection. We detect barcodes by barcode line extraction. We take the orientation distribution of barcode lines into consideration and define an orientation accumulator to determine dominant orientation of each cluster of barcodes. Once the orientation rectification of every barcode has been done, the gradient difference map can be calculated to highlight barcode area, and the highlighted areas are used to accelerate barcode detection. One important feature of our proposed method is that it uses the result of barcode line ex-traction to deal with the variation of size and direction of barcode lines, and increase the practicability and robustness of the method.