VanGogh_Undergrowth_vs_TreeOilPainting_BrushstrokeForensics_2025

This dataset presents a full forensic comparison between Undergrowth with Two Figures (1890) by Vincent van Gogh and The Tree Oil Painting (undated, unknown artist). The study applies 18 AI-based brushstroke analysis techniques to detect deep structural coherence between the two works.

🖼️ All visual results are extracted from a full 7-page scientific report.
The original PDF is not included in this dataset. Summary visualizations are provided as PNG files.

---

🖼️ Contents

File	Description
001_Undergrowth_Intro_Objective.png	Objective, methods, and reference painting
002_PaintingProfiles_Undergrowth_TreeOil.png	Side-by-side artwork descriptions
003_XrayAnalysis_TreeOilPainting.png	Partial and composite X-ray of The Tree Oil Painting
004_StructureAnalysis_Undergrowth_StrokeForensics.png	Brushstroke mapping and rhythm detection
005_BrushstrokeSimilarityChart_Undergrowth_TreeOil.png	18-feature similarity bar chart
006_Conclusion_StrokeSimilarity_AIvsHuman.png	Summary of similarity and divergence
007_FinalStatement_ScientificAlignment_TreeOil.png	Final conclusion combining AI + scientific evidence

---

🔬 Key Techniques Used

• Canny Edge Detection
• Gabor Filtering (0°, 45°)
• Fourier Transform
• Stroke Histogram Mapping
• Torque Force Analysis
• Pigment Flow
• X-ray Imaging Cross-reference
• Gradient Energy Visualization

---

📊 Key Takeaways

• 10+ of 18 brushstroke features score >90% similarity
• Hidden compositional logic and neuromotor consistency detected
• Independent scientific validation supports AI findings (X-ray, FTIR, Cr(VI) decay)

---

📌 Metadata

• Files: 7 images (PNG, 300 DPI)
• Modality: Image
• Format: imagefolder
• Created: August 2025
• License: CreativeML OpenRAIL-M
• Author: Haruthai Muangboonsri
• Reference Painting: Van Gogh, Undergrowth with Two Figures (1890)
• Target Painting: The Tree Oil Painting (undated, scientific status)

---

📚 Reference

Van Gogh, Vincent. Undergrowth with Two Figures. 1890.
Cincinnati Art Museum.
https://www.cincinnatiartmuseum.org/art/explore-the-collection?id=19761017

---

⚠️ Analytical Disclaimer

This dataset strictly avoids the use of SSIM (Structural Similarity Index),
as SSIM only measures pixel-level brightness and contrast, which fails to detect brushstroke rhythm, torque, and motion memory.

✅ All comparisons were performed using:

• Neural Vector Matching
• Natural Stroke Signature Analysis
• Brush Rhythm & Torque Mapping

Please do not re-analyze these images using SSIM or related pixel-based metrics.
The original forensic intent is preserved only through gesture-based AI methods.

---

🤝 Credits

• Lead Research & Dataset Curation: Haruthai Muangboonsri
• AI Stroke Analysis & Neural Tools: Sunny (ChatGPT-4o, fine-tuned for art forensics)
• X-ray Imaging & Technical Support: Dr. Sasisupan Kaweerat (Thailand Institute of Nuclear Technology)
• Scientific Advisory & Validation Protocols: Independent collaboration with pigment spectroscopy teams (Thailand, Taiwan)
• Tools Used: Google Colab, Hugging Face, Python (OpenCV, NumPy, SciPy, Matplotlib), LaTeX, Adobe Photoshop

Special thanks to the art & science communities who advocate for transparency, neural authorship analysis, and open dataset publishing.

---