Real-time implementation of two-dimensional barcode decoder for PDF417

This paper reports a method to extract the bar-space patterns directly from the gray-level PDF417 two-dimensional barcode image instead of binarizing it. Our method employs the location and the distance between extreme points of profile of each row or column of the barcode image. It proves to be very robust from the high convolutional distortion environments such as defocussing and warping, even under badly illuminating condition. The proposed algorithm is applied to implement a two-dimensional barcode decoder in real-time, which shows an excellent performance.