Advanced PDF Content Extractor#

A sophisticated Python tool for extracting tables and figures from complex, multi-layered PDF documents using state-of-the-art extraction libraries.

Features#

🔍 Layer-Aware Extraction: Handles complex PDFs with multiple layers
📋 Multi-Method Table Extraction: Uses Tabula, Camelot, and pdfplumber
🖼️ Advanced Figure Extraction: Vector graphics (SVG) + high-res rendering
📊 Rich Reporting: Markdown report with image previews and table samples
🤖 Smart Detection: Heuristic analysis to identify tables vs figures
⚡ CLI Interface: Easy-to-use command line interface

Installation#

Requirements#

  • Python 3.12+

  • Java Runtime Environment (for Tabula)

Setup#

# 1. Create virtual environment
python3.12 -m venv pdf_extraction_env

# 2. Activate environment
source pdf_extraction_env/bin/activate  # Linux/Mac
# or
pdf_extraction_env\Scripts\activate     # Windows

# 3. Install dependencies from requirements file
pip install -r requirements-pdfextractor.txt

# Alternative: Manual installation
pip install tabula-py camelot-py[cv] PyMuPDF pdfplumber pandas numpy Pillow opencv-python

Usage#

Basic Usage#

# Extract from PDF to default directory (extracted_content_FILENAME)
python advanced_pdf_extractor.py document.pdf

# Extract with verbose output
python advanced_pdf_extractor.py PDF_Proof.PDF --verbose

# Extract to custom directory
python advanced_pdf_extractor.py report.pdf --output my_extraction

Command Line Options#

usage: advanced_pdf_extractor.py [-h] [-o OUTPUT_DIR] [-v] [--version] pdf_file

positional arguments:
  pdf_file              Path to the PDF file to extract content from

options:
  -h, --help            Show help message
  -o OUTPUT_DIR         Output directory (default: extracted_content_FILENAME)
  -v, --verbose         Enable verbose output
  --version             Show version number

Output Structure#

The tool creates a comprehensive directory structure:

extracted_content_FILENAME/
├── tables/                 # CSV files with extracted table data
│   ├── tabula_lattice_table_1.csv
│   ├── camelot_stream_table_2.csv
│   └── pdfplumber_page_5_table_1.csv
├── figures/                # PNG and SVG files with extracted figures
│   ├── figure_page_1_highres.png    # High-resolution renders
│   ├── figure_page_2_vector.svg     # Vector graphics
│   └── embedded_page_3_img_1.png    # Embedded images
├── raw_data/               # Intermediate processing files
├── EXTRACTION_REPORT.md    # Comprehensive report with previews
└── extraction_report.json  # Machine-readable results

Extraction Methods#

Table Extraction#

  • Tabula: Best for form-based tables and structured data

  • Camelot: Excellent for complex layouts and scientific papers

  • pdfplumber: Precise text extraction and simple tables

Figure Extraction#

  • SVG Vector: Extracts vector graphics from all PDF layers

  • High-Resolution Render: 3x scaling for figure-rich pages

  • Embedded Images: Extracts actual embedded image files

Report Features#

The generated EXTRACTION_REPORT.md includes:

  • 📊 Summary statistics and method performance

  • 🖼️ Image gallery with PNG previews

  • 📋 Table previews showing first 10 rows as Markdown tables

  • 📁 Complete file listing with paths and metadata

Examples#

Example 1: Academic Paper#

python advanced_pdf_extractor.py research_paper.pdf --verbose

Output: extracted_content_research_paper/ with tables and figures

Example 2: Custom Directory#

python advanced_pdf_extractor.py complex_report.pdf -o report_extraction

Output: report_extraction/ with extracted content

Example 3: Processing PDF_Proof.PDF#

python advanced_pdf_extractor.py PDF_Proof.PDF

Result:

  • 📋 126 tables extracted (65 Tabula + 55 Camelot + 6 pdfplumber)

  • 🖼️ 67 figures extracted (46 vector + 19 high-res + 2 embedded)

  • 📁 Output in extracted_content_PDF_Proof/

Troubleshooting#

Common Issues#

  1. Missing Dependencies: The tool gracefully handles missing packages

  2. Java not found: Install OpenJDK 11+ for Tabula support

  3. Font warnings: Normal for complex PDFs, extraction continues

  4. Memory usage: Large PDFs may require more RAM

Dependency Management#

The tool is designed to show help and version information even without dependencies:

# These work without any packages installed:
python advanced_pdf_extractor.py --help     # ✅ Always works
python advanced_pdf_extractor.py --version  # ✅ Always works

# This shows helpful error if dependencies missing:
python advanced_pdf_extractor.py document.pdf
# ❌ Error: Required dependency not found: No module named 'pandas'
# 📦 To install all required dependencies, run:
# pip install -r requirements-pdfextractor.txt

Font Warnings#

Font warnings like “Start marker missing” are common with academic PDFs and don’t affect extraction quality.

Performance#

PDF_Proof.PDF Results (46 pages):

  • ⏱️ Processing time: ~2 minutes

  • 📊 Success rate: 126 tables + 67 figures extracted

  • 🎯 Methods used: All extraction methods successfully applied

  • 💾 Output size: ~15MB (high-resolution images included)

Technical Details#

Dependencies#

  • tabula-py: Java-based table extraction

  • camelot-py[cv]: Computer vision table detection

  • PyMuPDF: PDF manipulation and rendering

  • pdfplumber: Text-based PDF analysis

  • pandas: Data manipulation

  • numpy: Numerical operations

  • Pillow: Image processing

  • opencv-python: Computer vision

File Naming Convention#

  • Tables: {method}_{type}_table_{id}.csv

  • Figures: figure_page_{page}_{type}.{ext}

  • Output: extracted_content_{clean_filename}/

License#

This tool is designed for academic and research use with complex PDF documents.

Version: 2.0
Python: 3.12+
Last Updated: August 2025