Axis-Aligned Document Dewarping

2025-07-20

AI Summary

Existing document dewarping methods predominantly rely on supervised regression, neglecting the intrinsic axial symmetry geometric prior inherent in documents. This work proposes a geometry-aware dewarping framework: during training, it incorporates axial-alignment geometric constraints and feature-line optimization; during inference, it employs a geometry-driven preprocessing strategy. Furthermore, we introduce the Axis-Aligned Distortion (AAD) metric—a human vision-aligned evaluation criterion—used jointly for model optimization. Our approach requires no additional annotations and significantly improves bending correction accuracy. It achieves state-of-the-art performance across multiple benchmark datasets, outperforming prior methods by 18.2%–34.5% in AAD score. These results validate both the effectiveness and generalizability of explicitly modeling geometric priors in document dewarping.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search
• Machine Learning: Learning with Manifolds

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Document dewarping is crucial for many applications. However, existing learning-based methods primarily rely on supervised regression with annotated data without leveraging the inherent geometric properties in physical documents to the dewarping process. Our key insight is that a well-dewarped document is characterized by transforming distorted feature lines into axis-aligned ones. This property aligns with the inherent axis-aligned nature of the discrete grid geometry in planar documents. In the training phase, we propose an axis-aligned geometric constraint to enhance document dewarping. In the inference phase, we propose an axis alignment preprocessing strategy to reduce the dewarping difficulty. In the evaluation phase, we introduce a new metric, Axis-Aligned Distortion (AAD), that not only incorporates geometric meaning and aligns with human visual perception but also demonstrates greater robustness. As a result, our method achieves SOTA results on multiple existing benchmarks and achieves 18.2%~34.5% improvements on the AAD metric.

Problem

• Improves document dewarping using axis-aligned geometric constraints
• Reduces dewarping difficulty with axis alignment preprocessing
• Introduces Axis-Aligned Distortion metric for robust evaluation

Innovation

• Axis-aligned geometric constraint for training
• Axis alignment preprocessing for inference
• New Axis-Aligned Distortion metric for evaluation