jingsongliujing/OnnxOCR
A lightweight OCR system using ONNXRuntime for fast inference across multilingual text, documents, tables, and license plates.
Not currently ranked — collecting fresh signals.
star history
This project provides an OCR pipeline based on ONNX models and ONNXRuntime, enabling deployment without deep learning training frameworks. It supports text detection, recognition, layout analysis, table recognition, and license plate recognition. The system recently added integration with Qwen3.5-2B ONNX for end-to-end information extraction workflows.
Frequently asked
- What is jingsongliujing/OnnxOCR?
- A lightweight OCR system using ONNXRuntime for fast inference across multilingual text, documents, tables, and license plates.
- Is OnnxOCR open source?
- Yes — jingsongliujing/OnnxOCR is open source, released under the Apache-2.0 license.
- What language is OnnxOCR written in?
- jingsongliujing/OnnxOCR is primarily written in Python.
- How popular is OnnxOCR?
- jingsongliujing/OnnxOCR has 1.8k stars on GitHub.
- Where can I find OnnxOCR?
- jingsongliujing/OnnxOCR is on GitHub at https://github.com/jingsongliujing/OnnxOCR.