skills.rest
Deng-GuiFeng
Deng-GuiFengCOMMUNITY
Explore all skills by this author →
GitHub

extract-pdf-figure

Extract figures and tables from academic PDFs using bounding boxes.

3|Updated Feb 4, 2026
Search Tags:#pdf#tables#figures#figure-extraction#bounding-box#subfigures
One-click install
npx skills add https://github.com/Deng-GuiFeng/Paper2PPT --skill extract-pdf-figure
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: extract-pdf-figure
Source: https://github.com/Deng-GuiFeng/Paper2PPT/tree/main/.claude/skills/extract-pdf-figure
Command: npx skills add https://github.com/Deng-GuiFeng/Paper2PPT --skill extract-pdf-figure

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires PyMuPDF, Pillow, openai, python-dotenv, and includes scripts (resource) components.

What problem does it solve?

This Skill extracts figures and tables from academic PDFs with pixel-accurate bounding boxes using the Qwen3-VL model, enabling precise cropping of complete figures (title, legend, notes) or main content only, and supporting sub-figures like Figure 1(a).

Core Features & Use Cases

  • Grounding-based figure/table extraction with iterative quality assessment for high accuracy.
  • Full-content capture (including captions, legends, notes) or main-content-only extractions.
  • Sub-figure and batch extraction support, with high-DPI outputs for publication-quality figures.
  • Ideal for researchers, educators, and analysts who need to isolate visuals from papers or presentations.

Quick Start

Run the tool to extract a specific figure from a PDF by name, for example 'Figure 1' from PDFs/paper.pdf.

Frequently Asked Questions about extract-pdf-figure

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract figures and tables from an academic PDF?▼

To extract sub-figures from research papers, this tool uses an AI-driven bounding-box locator to identify and crop specific visual elements like Figure 1(a). It applies grounding-based extraction with iterative quality assessment to ensure precise sub-figure isolation.

Do I need an OpenAI API key to extract images from a PDF?▼

Yes, you need a configured OpenAI API key to extract images from a PDF. The tool relies on the Qwen3-VL model via the API to accurately locate bounding boxes for figures and tables, requiring PyMuPDF and Pillow as well.

Can I capture full captions and legends when extracting tables from PDFs?▼

Yes, you can capture full captions, legends, and notes when extracting tables from PDFs. The tool allows you to choose between full-content capture or main-content-only extraction for your cropped visuals.

What's the best way to crop publication-quality figures from research papers?▼

The best way to crop publication-quality figures from research papers is using a grounding-based extraction tool that supports high-DPI outputs. This approach ensures pixel-accurate bounding boxes for visually isolated figures.

Does this PDF figure extraction method work with batch processing?▼

Yes, this PDF figure extraction method works with batch processing, allowing you to extract multiple figures or tables from a single document. It iteratively assesses extraction quality to maintain high accuracy across all isolated visuals.

Related Skills

Explore More Education & Research Skills→
microsoftmicrosoft
2.2k

paper2assets

Extract text, figures, and metadata from PDF research papers into structured assets.

Official
Advanced
shushuznshushuzn
4

matplotlib

Generate publication-ready plots from scientific data using Matplotlib.

Community
Intermediate
BBufBBuf
721

vllm-qwen-vlm-omni-asr-optimization

Audit PR-driven optimizations for Qwen multimodal runtimes in vLLM.

Community
Advanced
gerald-icagerald-ica

pdf

Extract text and tables from PDFs and fill forms using Python libraries.

Community
Advanced