Aadhaar Masking Engine (aadhaar-masking)

A high-performance, color-preserving library for detecting and masking 12-digit Aadhaar card numbers in images (JPG, PNG, TIFF, WEBP) and multi-page PDF documents with C++ multi-threading acceleration and automatic Python ThreadPool fallback.

Features

• Color Preservation: CLAHE red-channel background filtering is used only for text detection. The black masking box is applied directly to the original image, preserving colors, notebook lines, and ink clarity.
• Multi-Page PDF Processing: Automatically splits multi-page PDFs, masks pages containing valid 12-digit Aadhaar numbers (with Verhoeff checksum validation), and recombines them into a single PDF.
• Hybrid Multi-Threading Engine:
  • Uses native C++ std::thread execution pool via pybind11 when a C++ compiler is available.
  • Automatically falls back to Python ThreadPoolExecutor when C++ binary extension is unavailable.
• Temporary File Cleanup: Intermediate OCR preprocessed files are cleaned up immediately after coordinate extraction.

Installation

pip install aadhaar-masking

Quick Start Usage

Python API

from aadhaar_masking import mask_aadhaar

# Mask an image file (JPG, PNG, TIFF, etc.)
mask_aadhaar("input_aadhaar.jpg", "masked_aadhaar.jpg")

# Mask a multi-page PDF document
mask_aadhaar("input_document.pdf", "masked_document.pdf")

Command Line Interface (CLI)

# Mask an image
aadhaar-masker input.jpg output_masked.jpg

# Mask a multi-page PDF
aadhaar-masker input_document.pdf output_masked_document.pdf

# Specify worker threads
aadhaar-masker input_document.pdf output_masked_document.pdf --threads 8

License

MIT License.