# ndl-lab/ndlocr-lite

NDLOCR-Liteアプリケーションのリポジトリ（ソースコードを含む）NDLOCR‑Lite application repository (including source code)

Repository: https://github.com/ndl-lab/ndlocr-lite
Canonical: https://ross.abutalabs.com/products/ndlocr-lite
Language: Python
License: CC-BY-4.0
License Family: other
Last push: 2026-08-14T01:40:42+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 97, release rhythm 87, longevity 14
- inputs: {"age_days": 201, "days_push": 20, "days_rel": 85, "gap_med": 4, "n_releases_24m": 8}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1309, forks 86 (observed 2026-08-28T04:04:19.414329+00:00)

## What it is
NDLOCR-Lite is a lightweight Japanese OCR application developed by the National Diet Library that converts digitized images of books and magazines into text. It runs on ordinary consumer computers without requiring a GPU, and is available both as a desktop GUI application and a Python command-line tool.

## Use cases
- extract text from scanned book images
- run OCR on Japanese documents without a GPU
- batch convert a directory of digitized page images to text
- OCR a single image file from the command line
- capture screen regions and convert them to text directly
- output OCR results as JSON or XML
- improve recognition of vertical Japanese text and tate-chu-yoko

## When to choose
- you need Japanese OCR on a laptop or desktop without a GPU
- you want a ready-to-use desktop app for Windows, Mac, or Linux
- you need to batch-process digitized library or archive images
- you want both GUI and CLI interfaces for OCR

## When to avoid
- you need high-accuracy OCR for non-Japanese languages
- you require GPU-accelerated large-scale OCR pipelines
- your file paths contain full-width Japanese characters when using the desktop app
- you need OCR integrated into a web service

## Facets
- artifact type: application
- maturity: active
- function: ocr, image-processing, nlp, cli, gui
- domain: computer-vision, files, pdf, developer-tools
- platform: windows, python, cross-platform, cli
- tags: japanese-ocr, text-recognition, digitized-documents, cpu-only, national-diet-library, handwriting-recognition, batch-processing, natural-language-processing, macos, linux, desktop

## Member repositories
- ndl-lab/ndlocr-lite (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:19.414329+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:50:00.295986+00:00, confidence not recorded.
  - readme: https://github.com/ndl-lab/ndlocr-lite (fetched 2026-08-28T04:04:19.414329+00:00, sha a05a32b245ff)
- Data as of 2026-08-30T08:39:29.467469+00:00.