# Error Patterns in Historical OCR: A Comparative Analysis of TrOCR and a Vision-Language Model

arXiv: https://arxiv.org/abs/2602.14524 · HTML: https://mlsanity.com/paper/2602.14524/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 21 references that are in the map.
- First posted on arXiv 16 Feb 2026.

## Builds on (most cited first, 8 of 8)

- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [OCRBench: On the Hidden Mystery of OCR in Large Multimodal Models](https://mlsanity.com/paper/2305.07895/) (2023)
- [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://mlsanity.com/paper/2109.10282/) (2021)
- [UReader: Universal OCR-free Visually-situated Language Understanding with Multimodal Large Language Model](https://mlsanity.com/paper/2310.05126/) (2023)
- [Ground Truth for training OCR engines on historical documents in German Fraktur and Early Modern Latin](https://mlsanity.com/paper/1809.05501/) (2018)
- [OCR4all -- An Open-Source Tool Providing a (Semi-)Automatic OCR Workflow for Historical Printings](https://mlsanity.com/paper/1909.04032/) (2019)
- [Lacuna Reconstruction: Self-supervised Pre-training for Low-Resource Historical Document Transcription](https://mlsanity.com/paper/2112.08692/) (2021)
- [Optimization of Image Processing Algorithms for Character Recognition in Cultural Typewritten Documents](https://mlsanity.com/paper/2311.15740/) (2023)

## Built on by (most cited first, 2 of 2)

- [End-to-End Text Line Detection and Ordering](https://mlsanity.com/paper/2606.04166/) (2026)
- [Rate-Utility Frontiers for Language Encodings: Comparing Tokens, Bytes, and Pixels Under Controlled Linguistic Content](https://mlsanity.com/paper/2607.16117/) (2026)

## Abstract (arXiv)

Optical Character Recognition (OCR) of eighteenth-century printed texts remains challenging due to degraded print quality, archaic glyphs, and non-standardized orthography. Although transformer-based OCR systems and Vision-Language Models (VLMs) achieve strong aggregate accuracy, metrics such as Character Error Rate (CER) and Word Error Rate (WER) provide limited insight into their reliability for scholarly use. We compare a dedicated OCR transformer (TrOCR) and a general-purpose Vision-Language Model (Qwen) on line-level historical English texts using length-weighted accuracy metrics and hypothesis driven error analysis. While Qwen achieves lower CER/WER and greater robustness to degraded input, it exhibits selective linguistic regularization and orthographic normalization that may silently alter historically meaningful forms. TrOCR preserves orthographic fidelity more consistently but is more prone to cascading error propagation. Our findings show that architectural inductive biases shape OCR error structure in systematic ways. Models with similar aggregate accuracy can differ substantially in error locality, detectability, and downstream scholarly risk, underscoring the need for architecture-aware evaluation in historical digitization workflows.