Study On The Algorithm Of Position And Recognition Of Tow-dimensional Bar Code

In our country, along with the rapid development of modern informationtechnology, two-dimensional bar code has been widely used in all ways of life. Theinformation of2-D bar code can be expressed at the same time in the horizontal andvertical directions, so capacity of the data is very large. And because thesecharacteristics, such as high information density, strong reliability, strong errorcorrection ability, variety information of expressions, in the field of automaticidentification also plays more and more important role. The2-D bar code application,using regular intake device capturing the two-dimensional bar code and then quicklyidentified, is an important development direction.In the real application, collected two-dimensional bar code images must havemany problems, such as the complex and diverse light, the complex shooting angle,difficulty to distinguish background patterns, the fouling bar code, and bar code beingfolded. This article in view of the above questions, based on image processing of therelevant knowledge, discusses the two-dimensional bar code location and recognitionalgorithm, which is the applicable in real life.Based on this article, the main content and results include the following:First of all, the bar code image area is being located on. This paper presents a barcode positioning algorithm based on SUSAN detection operator. Using this method,the complex background interference can be ignored. Then the two-dimensional barcode location can be achieved Secondly, the barcode image is being recognized. This paper used a method basedon the improved Hough transform algorithm for line detection, gets the information ofedge line segments and endpoint which is special search pattern of Chinese-sensiblecode, and calculates the angle of rotation, then the Chinese-sensible code specialsearch graph can be rotated to the lower left corner.Finally, based on the MATLAB environment, Chinese-sensible codeidentification system is being designed. And a each module algorithm can be realized,including the gray of, the location of bar code region, filtering and enhancing, the edgesegment detection, rotation correction, and grid sampling. After testing, results showthat the algorithm can effectively locate and identify code.