# faustomorales/keras-ocr

A packaged and flexible version of the CRAFT text detector and Keras CRNN recognition model.

Repository: https://github.com/faustomorales/keras-ocr
Canonical: https://ross.abutalabs.com/products/keras-ocr
Homepage: https://keras-ocr.readthedocs.io/
Language: Python
License: MIT
License Family: permissive
Topics: text-detection, keras-crnn, keras, ocr
Last push: 2025-09-22T13:16:10+00:00

## Health v2 (maintenance only)
Score: 42/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 43, release rhythm 8, longevity 100
- inputs: {"age_days": 2539, "days_push": 345, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1473, forks 371 (observed 2026-08-28T04:04:49.622007+00:00)

## What it is
A Python library packaging the CRAFT text detector and a Keras CRNN text recognition model into a high-level OCR pipeline. It supports pretrained inference and training custom detection/recognition models with TensorFlow 2.

## Use cases
- extract text from images in python
- detect and recognize text in photos
- run ocr locally without cloud apis
- train a custom text detection model
- find word bounding boxes in images
- compare ocr accuracy against cloud services

## When to choose
- you want a self-hosted, pip-installable OCR pipeline with pretrained weights
- you need both text detection and recognition in one API
- you want to fine-tune or train custom OCR models in Keras/TensorFlow

## When to avoid
- you need state-of-the-art accuracy on difficult text (newer models like PaddleOCR or cloud APIs may be better)
- you work in PyTorch rather than TensorFlow/Keras
- you need handwriting or non-Latin script support out of the box

## Facets
- artifact type: library
- maturity: maintenance
- function: ocr, computer-vision, machine-learning, image-processing
- domain: computer-vision, image-processing, machine-learning
- platform: python, cross-platform
- tags: text-detection, text-recognition, keras, tensorflow, craft, crnn, deep-learning, natural-language-processing

## Member repositories
- faustomorales/keras-ocr (main) score 42

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:49.622007+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 2026-08-30T04:34:37.479231+00:00, confidence not recorded.
  - readme: https://github.com/faustomorales/keras-ocr (fetched 2026-08-28T04:04:49.622007+00:00, sha 2aa4bfd272e1)
  - registry_pypi: https://pypi.org/pypi/keras-ocr/json (fetched 2026-08-29T11:42:33.894852+00:00, sha b1cfdda037a4)
- Data as of 2026-08-30T08:39:29.467469+00:00.