Optimizing dotted Arabic expiration date recognition with ARABEX: a convolutional autoencoder with bidirectional LSTM and CRNN approach

In this study, we introduced an approach for Automated Dotted Arabic Expiration Date Extraction using an Optimized Convolutional Autoencoder with a bidirectional LSTM. This approach was used to translate the Arabic dot matrix expiration dates into their corresponding filled-in dates. A custom lightw...

Full description

Saved in:
Bibliographic Details
Published in:International journal on document analysis and recognition Vol. 28; no. 4; pp. 555 - 572
Main Authors: Zaki, Hozaifa, Soliman, Ghada
Format: Journal Article
Language:English
Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2025
Springer Nature B.V
Subjects:
Accuracy
Computer Science
Datasets
Deep learning
Expiration
Image Processing and Computer Vision
Literature reviews
Neural networks
Original Paper
Pattern Recognition
Production lines
Recurrent neural networks
Synthetic data
ISSN:1433-2833, 1433-2825
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this study, we introduced an approach for Automated Dotted Arabic Expiration Date Extraction using an Optimized Convolutional Autoencoder with a bidirectional LSTM. This approach was used to translate the Arabic dot matrix expiration dates into their corresponding filled-in dates. A custom lightweight Convolutional Recurrent Neural Network (CRNN) model was then employed to extract the expiration dates. Owing to the lack of available dataset images for the Arabic dot matrix expiration date, we generated synthetic images by creating an Arabic dot matrix True Type Font matrix to address this limitation. Our model was trained on a realistic synthetic dataset of 3287 images, covering the period from 2019 to 2027, represented in the yyyy/mm/dd format. We then trained our custom CRNN model using the generated synthetic images to assess the performance of our model (ARABEX) by extracting the expiration dates from the translated images. Our proposed approach achieved a recognition accuracy of 99.4% on the test dataset of 658 images, while also achieving a Structural Similarity Index of 0.46 for image translation on our dataset and a score of 0.899 in different domain benchmark dataset. Our approach leverages significant improvements in efficiency and accuracy for Arabic dot matrix expiration date recognition, making it a well-suited solution for manufacturers aiming to enhance their production line efficiency and accuracy in handling expiration dates.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1433-2833
1433-2825
DOI:10.1007/s10032-024-00510-w