MatteViT: High-Frequency-Aware Document Shadow Removal with Shadow Matte Guidance

TL;DR

MatteViT is a novel transformer-based framework for document shadow removal that preserves high-frequency details using frequency amplification and shadow matte guidance, achieving state-of-the-art results.

Contribution

The paper introduces a high-frequency amplification module and a luminance-based shadow matte for improved shadow removal and detail preservation in documents.

Findings

Achieves state-of-the-art performance on RDD and Kligler benchmarks.

Better preserves text details, improving OCR accuracy.

Demonstrates robustness in real-world shadow removal scenarios.

Abstract

Document shadow removal is essential for enhancing the clarity of digitized documents. Preserving high-frequency details (e.g., text edges and lines) is critical in this process because shadows often obscure or distort fine structures. This paper proposes a matte vision transformer (MatteViT), a novel shadow removal framework that applies spatial and frequency-domain information to eliminate shadows while preserving fine-grained structural details. To effectively retain these details, we employ two preservation strategies. First, our method introduces a lightweight high-frequency amplification module (HFAM) that decomposes and adaptively amplifies high-frequency components. Second, we present a continuous luminance-based shadow matte, generated using a custom-built matte dataset and shadow matte generator, which provides precise spatial guidance from the earliest processing stage. These strategies enable the model to accurately identify fine-grained regions and restore them with high fidelity. Extensive experiments on public benchmarks (RDD and Kligler) demonstrate that MatteViT achieves state-of-the-art performance, providing a robust and practical solution for real-world document shadow removal. Furthermore, the proposed method better preserves text-level details in downstream tasks, such as optical character recognition, improving recognition performance over prior methods.