Analysis updated 2026-08-08 · repo last pushed 2025-03-26
Digitize lecture notes and math formulas using the Pix2Text plugin.
Extract Chinese-English text from images on older laptops using RapidOCR.
Handle multi-language document scanning with Tesseract and its language packs.
Recognize Chinese, English, or Japanese text using the WeChat OCR plugin.
| hiroi-sora/umi-ocr_plugins | open-gigaai/giga-world-1 | pluviobyte/rnskill | |
|---|---|---|---|
| Stars | 607 | 606 | 613 |
| Language | Python | Python | Python |
| Last pushed | 2025-03-26 | — | — |
| Maintenance | Stale | — | — |
| Setup difficulty | easy | hard | easy |
| Complexity | 2/5 | 5/5 | 2/5 |
| Audience | general | researcher | writer |
Figures from each repo's GitHub metadata at analysis time.
Download a plugin, unzip it, and drop it into a specific folder inside your Umi-OCR installation, no complex configuration needed.
This repository is a plugin library for Umi-OCR, an open-source OCR (optical character recognition) tool. In plain terms, Umi-OCR lets you extract text from images, screenshots, scanned documents, photos, entirely on your own computer, without sending anything to the cloud. This repo collects optional add-on engines that expand what the main software can recognize and how it performs. The way it works is straightforward: you download a plugin package, unzip it, and drop it into a specific folder inside your Umi-OCR installation. The software then picks up the new engine and lets you choose it. Each plugin is a different OCR "brain" with its own strengths. For example, PaddleOCR offers high accuracy and speed on capable machines, while RapidOCR is a lighter alternative that runs better on older or lower-spec hardware. There's also Pix2Text for mixed Chinese/English/math-formula content, Tesseract for multi-language support with strong English results, and a plugin that calls the Mistral AI cloud API for recognition without local processing. The target audience is anyone who needs to pull readable text from images but wants control over which engine does the work. A student digitizing lecture notes with math formulas might pick Pix2Text. Someone with an older laptop who just needs basic Chinese-English recognition might go with RapidOCR. A developer or business handling documents in many languages could use Tesseract and its downloadable language packs. There's even a plugin that taps into WeChat's built-in OCR for Chinese, English, and Japanese text. One notable aspect is the tradeoff between local and cloud processing. Most plugins run entirely on your CPU, meaning your data never leaves your machine, a plus for privacy, but performance depends heavily on your hardware. The Mistral AI plugin is the exception, offloading work to a cloud API, which removes hardware constraints but requires an internet connection and sends your images to a third party. This lets users pick the right balance of speed, accuracy, privacy, and hardware compatibility for their situation.
A collection of optional OCR engine plugins for Umi-OCR that let you extract text from images on your own computer, with different engines for different needs like speed, language support, or math formulas.
Mainly Python. The stack also includes Python, PaddleOCR, RapidOCR.
Stale — no commits in 1-2 years (last push 2025-03-26).
No license information was provided in the repository explanation, so usage rights are unclear.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.