A barcode is the representation of data including some information related to goods, offered for sale, which frequently appears in on-line fashion images. Detecting and decoding barcode has a variety of applications in the on-line marketplace. However, the existing method has limitation in detecting barcode in some backgrounds such as Tassels, strips, and texture in fashion images. So, our work focuses on identifying the barcode region and distinguishing a barcode from its patterns that are similar to it. We accomplish this by adding a post-processing technique after morphological operations. We also apply a Convolutional Neural Network (CNN) to solve this typical object detection problem. A comparison of the performance between our algorithm and a previous method will be given in our results. For decoding part, a package including current common types of decoding scheme is used in our work to decode the detected barcode. In addition, we add a pre-processing transformation step to process skewed barcode images in order to improve the probability of decoding success.

This work deals with localization and decoding of the linear barcode EAN-13 in the image. It describes types and qualities of barcodes and pose posibble accesses into the localization and decoding. The practical part concerns with the concept and implementation of the algorithm having regard to effectiveness for processing in real time. The localization is established on finding similar features of the barcode among the read rows of the image and decoding is made by finding of the supreme conformity towards referential code figures.