Recognition of PDF417 Barcode Based on Captured Images

Methods of recognizing PDF417 two-dimensional barcode captured by camera are discussed. An approach for localization of barcode according to gradient and morphological mathematics is proposed, that can detect the barcode from complex background correctly. Block rectification based on bilinear transformation is then given. For those non-linear distorted images, the rectification algorithm corrects geometric distortion effectively. A recognition algorithm of bar-space based on projective algorithm is proposed to recognize the symbol character, which solves the recognition-error problems caused by noise and illegibility. The result showed that the method con effectively recognize PDF417 two-dimensional barcode captured by camera.