• --folder_name: folder path
• --neighbor_box_distance: Nearest box distance
• --face_recognition: Face recognition method (dlib, ssd, haar)
• --rotation_interval: Id card rotation interval in degrees
• --ocr_method: ocr method (EasyOcr and TesseractOcr)

In Dlib and Haar face detection model, it is better to choose a rotation angle of less than 30 degrees, otherwise no face may be detected due to image inversion. Create a folder and put the ID card images in that folder

git clone git@github.com:musimab/Tc_ID_Card_OCR.git
mkdir images

python3 main.py --folder_name "images" --neighbor_box_distance 60 --face_recognition ssd --ocr_method EasyOcr --rotation_interval 60

python3 -m venv card_id_ocr_venv

source card_id_ocr_venv/bin/activate

pip3 install -r requirements.txt

The result image and cropped regions will be saved to ./outputs by default. The json data will be saved to ./test by default.

Finds all virtual envs in linux `locate -b '\activate' | grep "/home"``

1. deep learning based (Yolo SSD Faster Rcnn) identity card recognition model will be developed

Tc : 12345678909 Surname : MUSTAFA ALİ Name : YILMAZ DateofBirth : 07071999

The accuracy of the optical character system was evaluated according to 2 different criteria. The first of these is accuracy at the word level and the other is accuracy at the character level.

The evaluate.py function retrieves the predicted and actual values in json format

1. tc: 1303 / 1327 => 98.19 %
2. surname: 805 / 816 => 98.65 %
3. name: 742 / 746 => 99.46 %
4. dateofbirth: 976 / 976 => 100.0 %

1. tc : 0.96 %
2. surname : 0.91 %
3. name : 0.95 %
4. date: 1.0 %

https://github.com/sarra831/EasyOCR

If you want to support my work, you can buy me a coffee at https://www.buymeacoffee.com/mustafaboyuk