Large License Plate Detection Dataset

Introduction

The Large Registration Number Disclosure File is an extensive collection of images sourced from Google Open Images, designed for the development and validation of Registration number recognition algorithms. Dataset Structure

• • Training Set: Consists of 25,500 carefully selected images for algorithm training.
  • Test Set: Includes 400 images for final model evaluation after training.
• Labels: Each image subset is accompanied by corresponding YOLO format annotation files located in the “train,” “val,” and “test” directories within the “Labels” folder. These annotations detail the position and dimensions of license plates within the images, including class labels, bounding box coordinates, and object confidence levels.

Significance of the Dataset

With a total of 27,900 images, this File is among the largest of its kind, offering extensive data for training, validating, and testing Registration number Disclosure systems. 

Acknowledgments

This File incorporates images from Google Open Images and merges additional File from Roboflow, highlighting a cooperative effort to provide a substantial resource for Registration detection.

Invitation

We invite the Kaggle community to engage with this File, utilize it in developing innovative solutions, and contribute to the enhancement of Registration Number detection technology. This dataset aims to stimulate innovation and foster technological progress in the realm of automated license plate recognition.

This dataset is sourced from Kaggle.