Who wants Bangladesh National ID OCR extractor with 99.99% accuracy? #OCR #NID #bangladesh #Bangla

We just finished writing and testing a Bangladesh National ID data extractor with 100% accuracy. There are tons of potential use cases for this. As far as I know there is no such tool available. This software is written in python.

Sample API response from uploaded NID image in JSON:

{
  "text": "raw [...] text",  "photo": "...",  "signature": "...",
  "parsed": [
    "গণপ্রজাতন্ত্রী বাংলাদেশ সরকার",
    "Government of the People's Republic of Bangladesh",
    "NATIONAL ID CARD / জাতীয় পরিচয় পত্র",
    "নাম: মোঃ [...]",
    "Name: [...]",
    "পিতা: মোঃ [...]",
    "মাতা: [...]",
    "Date of Birth: [..] Mar 1990",
    "ID NO: 1990[...]"
  ],
  "id_no": "1990[...]",
  "dob": "[..] Mar 1990",
  "bn_mother": "[...]",
  "bn_father": "মোঃ [...]",
  "en_name": "Md. [...]",
  "bn_name": "মোঃ [...]"
}

These are the packages we used to achieve this:

• numpy
• tensorflow
• keras
• dash
• imutils
• pytesseract
• opencv
• xlrd
• pandas
• pdf2image
• dash-canvas

Here is the source image:

Here is the extracted text:

['নাম: রোমানা আক্তার মৌ', 'Name: Romana Akter Mou', 'মাতা: মোসা: আমেনা', 'Date of Birth: 19 Jan 1992', 'ID NO: 2610413965404']

Please reach us if you need this.