Inventor(s)

n/aFollow

Abstract

Illegible barcodes can hinder use of barcodes in many contexts. This document describes a machine learning (ML) model trained to read illegible barcodes. The training dataset used to train the ML model comprises illegible barcodes (e.g., having low contrast, small size, wear-and-tear) and corresponding groundtruth. The ML model is trained to generate a legible version of the barcode. During training, the model output is compared with the groundtruth and the difference is fed to a loss function to adjust the weights of the model. In field use, a user can capture a photo of an illegible barcode and access the trained ML model to obtain a legible version. The described ML model can improve barcode scanning in industrial settings, data centers, and other contexts where barcodes are used.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS