In the previous article, we validated and extracted text from the front side of the Aadhaar card, Now in this tutorial, we will see how we can validate and extract useful information from the backside of the Aadhaar card.

Importance of Aadhaar Card

An Aadhaar card is a unique number issued to every citizen in India and is a centralized and universal identification number. The Aadhaar card is a bio-metric document that stores an individual’s details in a government database and is fast becoming the government’s base for public welfare and citizen services.

So, In this article, we will focus on the verification and information extraction from the backside of the Aadhaar Card as it is also an important step in validating the document.

Every document has some unique features which makes it different from other documents. Distinguishable features of Aadhaar Card are -

1. State Emblem of India
2. GOI Symbol
3. QR Code
4. Aadhaar Logo
5. UIDAI Symbol

Identifying Features of an Aadhaar Card

Now, we have identified unique features so our next will be to train an Object Detection Model to identify these features and validate them. I will not go into much detail on how to train an object detection algorithm as I have covered it thoroughly in the last tutorial kindly refer to that :)

I have used TensorFlow 2 Object Detection API for training and validation purposes, you can use other deep learning frameworks like PyTorch, Keras depending on your requirements and tradeoffs.

#machine-learning #tensorflow #kyc #deep-learning #aadhaar