To Do - Project Improvement

[navchandar]

Features to be added for obfuscating images and documents, making it a better tool for testing AI text/object recognition systems:

1. Increase Randomness in Distortions:

   • Add more variability in askew, brightness, contrast, and sharpness settings to create a wider range of distortions.

2. Add Noise:

   • Implement a feature to add random noise to the images to simulate imperfections.

3. Overlay Patterns or Watermarks:

   • Add patterns or watermarks over the text to create additional visual noise.

4. Simulate Common Scanning Issues:

   • Introduce common scanning issues such as smudges, streaks, or partial text obfuscation.

5. Adjust Compression Levels:

   • Apply different levels of compression to the images to reduce quality and make text recognition more difficult.

6. Simulate Handwritten Annotations:

   • Overlay simulated handwritten annotations or marks on the document.

7. Color Variations:

   • Introduce variations in text and background colors to test AI's ability to recognize text in different color contrasts.

8. Blur and Focus Variations:

   • Apply varying levels of blur and focus to different parts of the document.

9. Rotation and Skew:

   • Randomly rotate or skew the document pages to different angles.

10. Metadata Preservation:

    • Preserve metadata, bookmarks, and links from the original PDF.

11. Support for Password-Protected PDFs:

    • Add functionality to handle password-protected PDFs.