Vietnamese ID Card Extractor v1.0

❗Newer version

Checkout ID Card Extractor v2.0

NOTICE

This repository is no longer actively maintained.

What does this mean? This repository is not being actively updated, and issues and pull requests may not be addressed. The code and resources provided may be outdated and might not be compatible with the latest software libraries and dependencies.

Why is it no longer maintained? There could be various reasons for discontinuing maintenance, including changes in project priorities, lack of resources, or the completion of the project's goals.

Can I still use the code or resources here? Certainly! You are welcome to use the code and resources in this repository as a reference or starting point for your own projects. However, please be aware that there may be better alternatives or updated versions available elsewhere.

Looking for alternatives? If you're seeking actively maintained alternatives or similar projects, consider searching on GitHub or other platforms, as there may be more up-to-date options to meet your needs.

Thank you for your interest in this repository, and we appreciate your understanding regarding its maintenance status.

Introduction

An ID Card extraction tool using Keras models which is implementation of YOLOv4 (Tensorflow backend) for detection and VietOCR for recognition.

Tutorial

I highly recommend you performing on virtual environment by using Anaconda.

Activating your virtual environment

conda activate [your_venv_name]

Cloning repository

git clone https://github.com/ntvuongg/vnese-id-extractor.git

Installing requirements

while read setup; do conda install --yes $setup || pip install $setup || conda install --yes -c  $setup $setup; done < setup.txt

Getting weights

You can click here to download my trained weights or get it from gdown module:

(You can pass below code if you have already installed gdown)

conda install gdown

gdown --id 1dAK75XvXP8L32FoSc1W3fFnt80Z32f1L

After downloaded weights, make sure you extract and put them in vnese-id-extractor/models/weights

Extracting your ID card

conda run -n [your_venv_name] --no-capture-output --live-stream python vnese-id-extractor/deploy.py

First time running this will take a lot of time. So, keep waiting :)

When server is ready, you can extract your ID card by accessing below:

http://127.0.0.1:5000/

You can drag and drop or browse your image to drag area.

After uploaded your image of ID Card, click Extract button to get information from ID card.