MRZScanner
The core functionality of this project is MRZ Text Recognition.
📄️ Introduction
MRZ (Machine Readable Zone) refers to a specific area on travel documents such as passports, visas, and identity cards, where information can be quickly read by machines. MRZ is designed and generated according to the guidelines in ICAO Document 9303, to speed up border control and improve the accuracy of information processing.
📄️ Installation
We provide installation through PyPI or by cloning the project from GitHub.
📄️ Quick Start
We provide a simple model inference interface that includes both preprocessing and postprocessing logic.
📄️ Advanced Settings
When calling the MRZScanner model, you can pass parameters to configure advanced settings.
📄️ Model Design
The world is simple, yet we complicate it ourselves.
📄️ Evaluation
We use two evaluation metrics to assess the performance of the model:
📄️ Dataset
Since MRZ images are mostly found in personal documents, we couldn’t find enough real training data.
📄️ Submission
In the real world, there's a wide variety of cases, and you might encounter situations where things don’t work as expected.
📄️ References
Researchs