GitHub·Code·Apache-2.0

Topdu/OpenOCR

OpenOCR: An Open-Source Toolkit for General-OCR Research and Applications, integrates a unified training and evaluation benchmark, commercial-grade OCR and Document Parsing systems, and faithful reproductions of the core implementations from a wide range of academic papers.

1.4k stars·Updated Jul 2026
chineseocrdocument-analysisdocument-parsingdocument-processingocrocr-pytorchscene-text-detectionscene-text-recognition
Open on GitHub

Linked papers

Code

UniRec-0.1B: Unified Text and Formula Recognition with 0.1B Parameters

arxiv · 2025

Why Lacuna links this resource
Confirmed by the paper and resource

The paper explicitly links to the GitHub repository as its official codebase and dataset. The repository is a multi-paper toolkit (OpenOCR) that explicitly lists the target paper ('UniRec-0.1B') under its own OCR algorithms and provides its implementation, making it a bundled resource.

Code

Decoder Pre-Training with only Text for Scene Text Recognition

ACM Multimedia · 2024

Why Lacuna links this resource
Confirmed by the paper and resource

The paper explicitly states that its code is available at the provided GitHub repository. The repository is a toolkit that integrates implementations from multiple academic papers, and it explicitly lists the target paper 'Decoder Pre-Training with only Text for Scene Text Recognition' along with a link to its specific configuration/code directory.

Code

Instruction-Guided Scene Text Recognition

arxiv · 2024

Why Lacuna links this resource
Confirmed by the paper and resource

The paper explicitly provides the GitHub repository URL as its code release. The repository is a toolkit for General-OCR research that bundles implementations for multiple papers, and it explicitly lists the target paper 'Instruction-Guided Scene Text Recognition' (IGTR) along with a link to its specific configuration/code within the repository.

Code

Out of Length Text Recognition with Sub-String Matching

arxiv · 2024

Why Lacuna links this resource
Confirmed by the paper and resource

The paper explicitly provides the GitHub repository URL as its official code release. The repository is a multi-paper toolkit (OpenOCR) that explicitly lists the target paper ('Out of Length Text Recognition with Sub-String Matching') and its method (SMTR) as one of the implemented models. Therefore, it is a bundled resource containing the paper's official code.

Code

SVTRv2: CTC Beats Encoder-Decoder Models in Scene Text Recognition

arxiv · 2024

Why Lacuna links this resource
Confirmed by the paper and resource

The paper explicitly provides the GitHub repository URL as its official code release. The repository is a multi-paper toolkit for OCR research that explicitly lists the target paper ('SVTRv2') and provides its models, configurations, and datasets, confirming it is a bundled resource containing the paper's implementation.