Abstract
EAN/UPC barcode is one of the most used barcode system in commodity production, logistics and warehouse operation. However it is unavoidable that the barcode will be damaged during the process of commodity transportation and sale. In this paper, a barcode recognition algorithm based on deep learning is proposed for the recognition of damaged barcode. In our proposal, a convolutional neural network is designed for barcode recognition. The CNN based on deep learning is basically constructed of six convolution layers and three full connected layers. A hundred thousand barcode images with simulated degradation are generated as dataset to train the model, and a custom loss function is utilized to boost the recognition performance. The experiment result shows that recognition rate is up to 99.43%.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
