COVID-Text-Extractor

Extract COVID related info from Images, PDFs and Texts using Google's Tesseract-OCR, PDFMiner.six and Regular Expressions.

How it can be used?

• For building APIs
• For creating bots (Telegram, Twitter...)
• For Automatting COVID Data extraction, and more!

USAGE:

python3 image_to_text.py <path_to_image_file>

For Extracting Text on a random image from test directory:

python3 image_to_text.py

Requirements:

• Python 3.9.4
• OpenCV (Contrib) [4.5.1.48]
• PyTesseract [0.3.7]
• Pillow [8.2.0]
• Numpy [1.20.2]

Installing requirements...

pip3 install -r requirements.txt

IMPORTANT Dependency:

Install Tesseract Training Data (english) according to your OS

For Arch Linux Users (mine is Manjaro):

pacman -S tesseract-data-eng

For Other Users:

Click here to Download eng.traineddata and configure the Python script accordingly.

What it Currently Extracts in Texts and Images:

• Oxygen (if Oxygen, Cylinder, Cans, Concentrator or Refill is in text)
• Verified (if source is verified)
• Plasma (if keyword is used in text)
• Email (all emails)
• Age (extracts age)
• Blood-Groups (All blood groups)
• Phone-Numbers (All phone numbers)
• Required (If Required keyword is used in text)
• Help (If Required keyword is used in text)
• Food (breakfast, lunch, dinner)
• Urgent (If urgent or required keyword is used in text)
• ICU-Beds (if ICU or Bed keyword is used in text)
• Ventilator (if Ventilator keyword is used in text)
• Ambulance (if Ambulance keyword is used in text)
• Without (if without keyword is used in text)
• Free (if Free keyword is used in text)
• Report (if Report keyword is used in text)
• Fabiflu (if Fabi-flu or Fabiflu is in text)
• Medicine (if Medicine in text)
• Vaccine (if vaccine or vaccination is in text)
• Remdesivir (if Remdesivir is in text)