sergiomsilva/alpr-unconstrained
A deep learning system for detecting and recognizing license plates in images using YOLO-based object detection and custom recognition models.
Velocity · 7d
+0.6
★ / day
Trend
→steady
star history
This repository implements a neural network pipeline for license plate detection and recognition in unconstrained scenarios. It combines Darknet/YOLO for the detection stage with a separate recognition stage, built using Keras and TensorFlow. The project accompanied an ECCV 2018 paper and provides trained models, training scripts, and evaluation tools for the ALPR task.