of Neural Network Approaches to Restoring Damaged QR Codes

In an increasingly digital world, Quick Response (QR) codes have become essential tools for fast and reliable access to information, bridging the gap between physical and digital environments. However, these codes are often exposed to external factors such as environmental degradation, mechanical da...

Full description

Saved in:
Bibliographic Details
Published in:International Conference on Actual Problems of Electronic Instrument Engineering proceedings pp. 1 - 6
Main Authors: Yusupova, Shohida, Faxriddin, Bekchanov, Mardonbek, Jumanazarov
Format: Conference Proceeding
Language:English
Published: IEEE 14.11.2025
Subjects:
Adaptation models
Biological system modeling
Computer architecture
computer vision
Convolutional neural networks
damaged QR codes
deep learning
digital restoration
Distortion
error correction
Generative adversarial networks
image processing
Image reconstruction
Image restoration
machine learning
Neural networks
QR code recognition
QR code restoration
QR codes
ISSN:2473-8573
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In an increasingly digital world, Quick Response (QR) codes have become essential tools for fast and reliable access to information, bridging the gap between physical and digital environments. However, these codes are often exposed to external factors such as environmental degradation, mechanical damage, printing errors, or incorrect scanning, which can significantly reduce their readability and reliability. The inability to decode damaged QR codes may lead to data loss, communication failures, or disruptions in commercial, industrial, and personal contexts. Restoring images affected by such distortions is therefore a critical problem in computer vision and image processing. Traditional error correction techniques embedded in QR codes provide a limited level of recovery and are ineffective in cases of severe damage. With the advancement of artificial intelligence, deep learning methods and neural networks have emerged as powerful solutions for addressing these challenges. Neural networks not only adapt to varying types of distortions but also learn complex structural patterns, enabling more accurate restoration. This paper presents a classification of neural network approaches to damaged QR code restoration, highlighting convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), autoencoders, and transfer learning models. Their characteristics, advantages, limitations, and performance indicators are analyzed in the context of real-world applications. The study emphasizes the growing role of neural architectures in ensuring robust, efficient, and reliable QR code recovery, while also outlining potential directions for future research and development in digital restoration technologies.
ISSN:2473-8573
DOI:10.1109/APEIE66761.2025.11289385