# githubharald/SimpleHTR

Handwritten Text Recognition (HTR) system implemented with TensorFlow.

Repository: https://github.com/githubharald/SimpleHTR
Canonical: https://ross.abutalabs.com/products/simplehtr
Homepage: https://harald-scheidl.medium.com/2326a3487cd5
Language: Python
License: MIT
License Family: permissive
Topics: handwritten-text-recognition, ocr, recurrent-neural-networks, tensorflow, deep-learning, machine-learning
Last push: 2026-06-25T16:14:06+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 89, release rhythm 35, longevity 100
- inputs: {"age_days": 3003, "days_push": 69, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2183, forks 916 (observed 2026-08-28T04:06:23.455601+00:00)

## What it is
A Handwritten Text Recognition (HTR) system implemented in TensorFlow that recognizes text from images of single words or text lines, trained on the IAM dataset. It includes pretrained models, training/validation/inference modes, and multiple CTC decoding options including word beam search.

## Use cases
- recognize handwritten text in images
- transcribe scanned handwritten words or lines
- train an HTR model on the IAM dataset
- experiment with CTC decoders for handwriting recognition
- build OCR for handwritten notes

## When to choose
- you need to recognize handwritten (not printed) text in word or line images
- you want a simple, well-documented TensorFlow HTR reference implementation
- you want pretrained models with reasonable accuracy out of the box

## When to avoid
- you need full-page document parsing rather than single words or lines (use HTRPipeline instead)
- you need production-grade OCR for printed text
- you need a maintained, actively developed solution

## Facets
- artifact type: library
- maturity: maintenance
- function: ocr, machine-learning, deep-learning, image-processing
- domain: computer-vision, machine-learning, deep-learning, image-processing
- platform: python, cross-platform
- tags: handwritten-text-recognition, tensorflow, ctc, rnn, iam-dataset, htr

## Member repositories
- githubharald/SimpleHTR (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:23.455601+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 2026-08-30T02:48:14.974245+00:00, confidence not recorded.
  - readme: https://github.com/githubharald/SimpleHTR (fetched 2026-08-28T04:06:23.455601+00:00, sha 632e4a8a3a5f)
- Data as of 2026-08-30T08:39:29.467469+00:00.