Barcode Searching And Recognition Algorithms For Good Images Based On Image Processing

Barcode technology is a automatic identification technology developed from the20th century on the computer and information science field, it is a comprehensive technology consist of barcode theory, photoelectricity technology, computer theory and information theory. Nowadays, barcode is becoming a global commonable business language and also a method of data standardization, achieving the uniform coding and identification of goods information. There are several advantages in barcode, such as fast input speed, high reliability, collecting large amount of information, portability, photocopiable property, long life and irreversible property.As developing fast and becoming mature, the digital image processing technology are playing an important role on barcode field. Firstly, by using digital image processing technology can utilize the whole barcode information, will not make mistake because of local loss. Secondly, each image pixel has256gray levels, and this precision can almost satisfy application requirements on any field. At last, each pixel has strong relation to others, so each pixel is not independent, that is also an advantage which common barcode processing not have.This paper primarily researches barcode location and recognition algorithms for goods image. Barcode location part firstly analyses the essential difference between barcode image and others, then scanning the whole images based on this characteristic and choosing area which satisfied this characteristic, so we can get several barcode area information. Secondly, the algorithm researches improvement by analyzing experimental result, and scanning this images the second time by using the result that the first time scanning gets.So this is a machine learning process.At last, filter the result we get by barcode spatial information and price card images characteristic, so that the barcode location is accomplished. Barcode recognition is based on barcode location, mainly aiming at EAN-13barcode in goods image. Firstly, make barcode images preprocessing by direction filter and image segmentation. Then recognise every character in code by similar edge distance method. At last, the algorithm imports verification and error correction of recognition result. After algorithm discussing, this paper design a system of barcode location and recognition in good images. This system contains image collecting, image transmission, barcode location and barcode recognition, and introducing the design scheme and working process. In this paper, resolution of barcode image is low and barcode is a small part of the whole image, moreover, there are so many fuzzy parts in image.So that, it is very difficult to not only barcode location but also recognition. The experimental results show that the accuracy rate of barcode location can reach more than95%,and the accuracy rate of barcode recognition can reach more than90%. And that the system working and image processing both costs a little time.