MFE-GAN: Efficient GAN-based Framework for Document Image Enhancement and Binarization with Multi-scale Feature Extraction

2025-12-16

AI Summary

To address the low training and inference efficiency caused by multi-channel independent GANs in degraded color document image enhancement and binarization, this paper proposes an efficient, lightweight end-to-end GAN framework. Our method introduces: (1) a multi-scale feature extraction module integrating Haar wavelet transform and normalization to explicitly model cross-scale texture and structural information; (2) parameter-efficient generator and discriminator architectures; and (3) a joint adversarial loss coupled with an adaptive reconstruction loss. Experiments on Benchmark, Nabuco, and CMATERdb datasets demonstrate that our approach significantly reduces total training and inference time while achieving competitive performance—comparable to state-of-the-art methods—in terms of PSNR, F-Score, and other quantitative metrics. The framework thus achieves an effective balance between computational efficiency and restoration accuracy.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Document image enhancement and binarization are commonly performed prior to document analysis and recognition tasks for improving the efficiency and accuracy of optical character recognition (OCR) systems. This is because directly recognizing text in degraded documents, particularly in color images, often results in unsatisfactory recognition performance. To address these issues, existing methods train independent generative adversarial networks (GANs) for different color channels to remove shadows and noise, which, in turn, facilitates efficient text information extraction. However, deploying multiple GANs results in long training and inference times. To reduce both training and inference times of document image enhancement and binarization models, we propose MFE-GAN, an efficient GAN-based framework with multi-scale feature extraction (MFE), which incorporates Haar wavelet transformation (HWT) and normalization to process document images before feeding them into GANs for training. In addition, we present novel generators, discriminators, and loss functions to improve the model's performance, and we conduct ablation studies to demonstrate their effectiveness. Experimental results on the Benchmark, Nabuco, and CMATERdb datasets demonstrate that the proposed MFE-GAN significantly reduces the total training and inference times while maintaining comparable performance with respect to state-of-the-art (SOTA) methods. The implementation of this work is available at https://ruiyangju.github.io/MFE-GAN.

Problem

• Enhances degraded document images for OCR
• Reduces training and inference time of GANs
• Improves binarization with multi-scale feature extraction

Innovation

• MFE-GAN framework with multi-scale feature extraction
• Haar wavelet transformation for preprocessing document images
• Novel generators, discriminators, and loss functions