JSS Mahavidyapeetha
JSS Science and Technology University, Mysuru – 570 006
DEPARTMENT OF COMPUTER APPLICATIONS
PDF Tamper Detection Using Metadata Analysis.
Guide : Presented By,
Misba Firdose – 02JST24PMC026
Mrs. Nandini S – Assistant Professor
Shreya Nayak – 02JST24PMC044
1
Department of Computer Applications , JSS STU
Problem Statement
• Rise in fraudulent documents.
• Manual verification is time-consuming and error-prone.
• Need for an automated system to detect document tampering efficiently.
2
Department of Computer Applications
Objective of the Project
• Develop a system to detect tampering in PDF documents across various domains.
• Detect unauthorized modifications in PDF documents.
• Verify authenticity using metadata.
• Compare structure and content for visual or hidden changes through structural
analysis.
• Ensure document integrity for official use.
3
Department of Computer Applications
Literature Review
SI Title Year Type Methodology
Enhanced Computer Vision Framework Image preprocessing with OpenCV, thresholding, contour
1 2024 Journal
for Tamper Detection in PAN Cards extraction, SSIM analysis, CNN model
Extraction and analysis of metadata from Office/OpenOffice
Disclosing Private Information from
2 2009 Conference docs ([Link], [Link], [Link]), examining hidden
Metadata, Hidden Info and Lost Data
elements, printers, user data, and document history
Block-based copy-move detection, Error Level Analysis (ELA),
Forgery Detection and Localization in
3 2020 Conference and metadata analysis with detection rules based on EXIF,
Scanned Documents
XMP, ICC, IPTC headers
Graph-based analysis of OCR-extracted characters using
OCR Graph Features for Manipulation
4 2020 arXiv Paper bounding boxes; features include height, width, alignment, Hu
Detection in Documents
moments, distance; classification via Random Forest
Summary of the Literature Survey
• A vision-based method detects PAN card tampering using SSIM and contour analysis.
• Metadata from Office/OpenOffice files reveals author info, edit history, and device
details.
• Scanned document forgery is identified using copy-move detection, ELA, and
metadata checks.
• OCR-based graph features capture alignment and size mismatches to spot character-
level edits.
Department of Computer Applications 5
Brief Description of Proposed Work
• Input: Original and suspect PDF documents.
• Process:
• Extract and compare PDF metadata which includes author, creation date, version.
• Perform layout structure analysis for visual differences.
• Use OCR to extract and match textual data.
• Output: Tamper alerts and discrepancy highlights.
6
Department of Computer Applications
Software Requirements
•Languages: Python
•Platform: Windows/Linux
7
Department of Computer Applications
Hardware Requirements
•Processor: Intel i5 or above
•RAM: Minimum 8 GB
•Storage: At least 100 GB free space
•Display: 1080p resolution or higher
8
Department of Computer Applications
THANK YOU
9
Department of Computer Applications