1. chineseocr_lite: Lightweight Chinese OCR Project
chineseocr_lite is a lightweight Chinese OCR project that provides the functionality to convert Chinese characters from images into text strings. It supports vertical text and inference using ncnn, mnn, and tnn frameworks (dbnet (1.8M) + crnn (2.5M) + anglenet (378KB)), with a total model size of only 4.7M.
2. Capture2Text: Portable OCR Tool
Capture2Text is a portable OCR tool that allows quick OCR on selected screen regions using keyboard shortcuts. By default, the generated text is saved to the clipboard. It supports over 90 languages including Chinese, English, French, German, Japanese, Korean, Russian, and Spanish. Capture2Text is portable and does not require installation.
3. Tesseract.js: JavaScript Library for Image Text Recognition
Tesseract.js is a JavaScript library that can recognize text in almost any language from images.
4. Tesseract OCR: Image Recognition Library
Tesseract is an open-source OCR engine developed by HP Labs and later maintained by Google. It is free, supports multiple languages, and works on various platforms.
5. chineseocr_lite: Lightweight Chinese OCR
chineseocr_lite is a lightweight Chinese OCR project that converts Chinese text from images into strings. It supports vertical text and inference using ncnn, mnn, tnn frameworks (dbnet (1.8M) + crnn (2.5M) + anglenet (378KB)), with a total model size of only 4.7M.
6. Textinator: macOS OCR Tool
Textinator is a simple macOS status bar/menu bar application that performs automatic text detection on screenshots.
7. MathOCR: Mathematical Formula Recognition System
MathOCR is a printed mathematical formula recognition system written in Java. It recognizes mathematical formulas from images and converts them into corresponding LaTeX code.
8. Shudong OCR: Cross-platform OCR Tool
Shudong OCR is a cross-platform OCR tool. The full version includes the runtime environment, while the simplified version may not work without the full version. It uses cloud-based OCR services, meaning an internet connection is required.
9. MMOCR: Text Detection Tool Based on PyTorch
MMOCR is an open-source toolbox based on PyTorch and mmdetection, focusing on text detection, text recognition, and downstream tasks such as key information extraction. It supports text detection, text recognition, and downstream tasks like key information extraction.
10. PaddleOCR: OCR Toolkit Based on PaddlePaddle
PaddleOCR is an OCR toolkit based on PaddlePaddle, containing an ultra-lightweight Chinese OCR model with a total size of only 8.6M. The model supports mixed Chinese-English numeral recognition, vertical text recognition, and long text recognition. It also supports various training algorithms for text detection and recognition.
11. Umi-OCR: Offline Text Recognition Tool
Umi-OCR is a free, open-source, batch-processing offline OCR software based on PaddleOCR, suitable for Windows 10/11. Umi-OCR allows batch image import and saves results to local txt/md/jsonl formats. It also supports instant screenshot recognition.
12. C-OCR: Ctrip OCR Project
C-OCR is an OCR project developed by Ctrip, mainly for the recognition of travel-related documents such as ID cards, passports, train tickets, and visas. The project consists of four parts: rejection, detection, recognition, and post-processing.
13. MindOCR: OCR Development and Application Toolbox
MindOCR is an open-source OCR development and application toolbox based on the MindSpore framework. It helps users train and apply industry-leading text detection and recognition models, such as DBNet/DBNet++ and CRNN/SVTR, to meet the needs of image text understanding.
14. dpScreenOCR: Lightweight OCR Tool
dpScreenOCR is a free, open-source program for recognizing text on the screen. Powered by Tesseract, it supports over 100 languages and can split independent text blocks, such as columns.
15. RapidOCR: Cross-platform OCR Library
RapidOCR is a cross-platform OCR library based on PaddleOCR, OnnxRuntime, and OpenVINO. It is fully open-source, free, and supports offline quick deployment. It is cross-platform and supports Chinese and English by default, with other languages requiring conversion.
