Back to projects

Aadhaar OCR & Information Extraction Platform

AadharLens

A document-processing platform that extracts structured Aadhaar information from front and back images using OCR, image preprocessing, and advanced parsing.

Overview

AadharLens handles the path from uploaded document imagery to validated structured data, using local and cloud OCR options with image preprocessing and parsing steps.

Key features

  • Local Tesseract.js OCR
  • AWS Textract OCR
  • Name extraction
  • DOB extraction
  • Gender extraction
  • UID extraction
  • Address extraction
  • Pincode extraction
  • Mobile number extraction
  • Regex validation
  • Zod validation
  • Front/back consistency checks

Architecture

OCR processing pipeline

  1. Image
  2. Sharp preprocessing
  3. OCR
  4. Parsing
  5. Validation
  6. Structured data

Engineering highlight

  • Local Tesseract.js
  • AWS Textract
  • Sharp preprocessing
  • Regex
  • Zod

Screenshots

AadharLens screenshot