sergiomsilva/alpr-unconstrained

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

1770 stars · 601 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

An implementation of the ECCV 2018 paper 'License Plate Detection and Recognition in Unconstrained Scenarios', combining a Darknet-based detector with Keras/TensorFlow recognition networks. It detects vehicle license plates in arbitrary images and outputs recognized plate text, with scripts for running inference and retraining the detector.

Use cases

• read license plates from photos automatically
• detect and recognize car plates in unconstrained images
• build an automatic license plate recognition pipeline
• fine-tune a plate detector on my own annotated dataset
• batch process a folder of vehicle images and export plate text to CSV

When to choose

• you need research-grade ALPR that works on full unconstrained scenes, not cropped plates
• you want to retrain or fine-tune the plate detector on custom data
• you can run on Linux with Python 2.7-era Keras/TensorFlow and optionally CUDA

When to avoid

• you need a maintained production system on modern TensorFlow/Python 3
• you only need plate recognition on pre-cropped plate images
• you require a permissive license - the license is non-standard
• you need Windows or macOS support out of the box

Facets

library · maturity maintenance

computer-vision image-processing ocr machine-learning deep-learning computer-vision image-processing artificial-intelligence python cpp c alpr license-plate-recognition darknet keras tensorflow yolo eccv-2018 research-code linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="sergiomsilva/alpr-unconstrained")