One-Dimension Barcode Localization Algorithm With Complex Background

With the continuous development of devices like mobile phone camera and scanner, getting image is becoming more and more convenient. At the same time, barcode has been widely used in many aspects like consumption, logistics, entertainment and payment. Considering the limitation of traditional barcode decoding method using laser scanning, it is being more and more necessary to get a new approach to decode barcode. With this trend, decode barcode by digital image processing is being much more important.With the digital image processing decoding approach, barcode localization must be performed before decoding. Better localization performance help ensure that the decoder can parse the barcode successfully. In this paper, we present a novel one-dimension barcode localization approach in complex background. Our algorithm is based on region-based image analysis and morphology processing. The sole of our algorithm is edge detection and orientation histogram. We compared our method with some published method from the literature by different scales and blurs. With better edge detection and candidate patch filter, our approach achieved better localization performance with millisecond in normal size image. Besides, experiment result show that our approach is insensitive to many factors like the barcode slope, geometric deformation, barcode position, barcode percentage in image and barcode count. In summary, our barcode localization algorithm can help decoder speed up the process in complex background.