期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2013
卷号：50
期号：3
出版社：Journal of Theoretical and Applied
摘要：PDF417 barcode identification is very sensitive to the skew angle while existing tilt angle detection algorithms are defective because of great calculation cost or high complexity or uncertain result. In this paper, a series of morphological transforms are applied to extract PDF417 out of complex background. Firstly, pre-process captured PDF417 images by noise removal and illumination equalization and then binarize it. Secondly, barcode�s skew angle is detected through measuring eroded area by dot pair or linear structure element based on the PDF417 barcode�s directional texture characteristics. Finally, reconstruction operation is utilized to obtain the full PDF417 barcode. The simulation shows that the algorithm can detect the PDF417 tilt angle exactly, and segment the barcode from complex noisy background.
