Study On Two Dimensional Barcode Recognition In Inartificial Conditions

The 2D barcode which emerged with the development of information technology, has a promising future. At the same time, the barcode reader runs to the fourth generation which is based on image processing. However, the current reader has many limitations. So this paper presents a barcode read system based on computer vision theory, it can not only improve the functions of the current reader, but also provide many image capture devices with the ability of reading the barcodes.This paper studies on the automatic recognition technology of 2D barcode captured under inartificial conditions and presents a reading system of 2D barcode according to the computer vision theory. The system studies the blurred images caused by losing focus and deformed images firstly and uses the preprocessing methods accordingly. Then image segmentation is processed to extract the barcode area. Additionally, if there has many barcodes in images, pattern analysis is presented more. Finally, the barcodes is recognized automatically. An estimate method of the optimal blur parameter is presented in this paper, the blurred images can be restored by the Wiener filter then. In the case of the deformed barcode images, this paper suggests a block rectification method, choosing the control points according to the barcode structure, then confirming the deformed blocks, rectifying every part later. Two image segmentation methods are also presented in this paper, self-adaptive threshold segmentation method and texture segmentation method based on the Gabor filter. When many barcodes occur in the same image, pattern analysis based on morphology is used further to divide every part. During the barcode analysis phase, projection means is used to extract each column, then the bar-space ratio sequence is used to located barcode precisely and to recognize thecharacter. Also the foreknown knowledge can be used to improve the recognition performance. In addition, this paper put forward a scheme based on image processing to meet 3D barcode captured in the complicated environment recognition problem. At last, the 2D barcode fabrication prevented certificate management system using barcode recognition technology is introduced.Researches show that the reading system of 2D barcode built on the computer vision theory can enhance the flexibility and reliability of reading. Without the auxiliary light, the auto-focus device and the limitations to the angle and the depth of field, the deformed, blurred and multi-barcodes, etc barcode images can be recognized automatically. So it is an applicable reading system with high performance. Also the 3D barcode reading system and the 2D barcode fabrication prevented certificate management system using barcode recognition technology are practical.