預測損壞歷史文件的原始外觀

Predicting the Original Appearance of Damaged Historical Documents

December 16, 2024

作者: Zhenhua Yang, Dezhi Peng, Yongxin Shi, Yuyi Zhang, Chongyu Liu, Lianwen Jin

cs.AI

摘要

歷史文獻包含豐富的文化寶藏，但隨著時間推移，常常遭受嚴重損壞，包括缺字、紙張損傷和墨跡侵蝕。然而，現有的文獻處理方法主要集中在二值化、增強等方面，忽略了對這些損壞的修復。為此，我們提出了一個新任務，稱為歷史文獻修復（HDR），旨在預測損壞歷史文獻的原始外觀。為填補這一領域的空白，我們提出了一個大規模數據集 HDR28K 和一個基於擴散的網絡 DiffHDR 用於歷史文獻修復。具體而言，HDR28K 包含 28,552 張損壞-修復圖像對，帶有字符級標註和多風格降解。此外，DiffHDR 通過添加語義和空間信息以及精心設計的字符感知損失，以實現上下文和視覺一致性，擴展了基本的擴散框架。實驗結果表明，使用 HDR28K 訓練的 DiffHDR明顯優於現有方法，在處理真實損壞文檔方面表現出色。值得注意的是，DiffHDR 還可以擴展到文檔編輯和文本區塊生成，展示了其高靈活性和泛化能力。我們相信這項研究可能開創文獻處理的新方向，並有助於珍貴文化和文明的傳承。數據集和代碼可在 https://github.com/yeungchenwa/HDR 找到。

English

Historical documents encompass a wealth of cultural treasures but suffer from severe damages including character missing, paper damage, and ink erosion over time. However, existing document processing methods primarily focus on binarization, enhancement, etc., neglecting the repair of these damages. To this end, we present a new task, termed Historical Document Repair (HDR), which aims to predict the original appearance of damaged historical documents. To fill the gap in this field, we propose a large-scale dataset HDR28K and a diffusion-based network DiffHDR for historical document repair. Specifically, HDR28K contains 28,552 damaged-repaired image pairs with character-level annotations and multi-style degradations. Moreover, DiffHDR augments the vanilla diffusion framework with semantic and spatial information and a meticulously designed character perceptual loss for contextual and visual coherence. Experimental results demonstrate that the proposed DiffHDR trained using HDR28K significantly surpasses existing approaches and exhibits remarkable performance in handling real damaged documents. Notably, DiffHDR can also be extended to document editing and text block generation, showcasing its high flexibility and generalization capacity. We believe this study could pioneer a new direction of document processing and contribute to the inheritance of invaluable cultures and civilizations. The dataset and code is available at https://github.com/yeungchenwa/HDR.