Analysis of Buddha Images Around the Phra Pathom Chedi Using Feature Matching

A computer vision project for identifying Buddha statues at Phra Pathom Chedi using image feature matching techniques.

---

🎯 Project Overview

This project focuses on the recognition of Buddha statues around Phra Pathom Chedi by detecting and matching image features. A dataset was created from real-world photos taken around the site, and green screen techniques were applied to standardize backgrounds and improve detection performance.

---

⚙️ Workflow

1. Data Collection

   • Photographs of Buddha statues were taken from various locations around Phra Pathom Chedi.

2. Image Preprocessing

   • Backgrounds were manually turned into green screen using Paint, Photoshop, or similar tools.

3. Metadata Annotation

   • A buddha_names.txt file maps image filenames to their corresponding Buddha statue names.

4. Feature Detection & Matching

   • Uses the following algorithms:
     • ORB
     • AKAZE
     • KAZE
     • SIFT
   • Input images are compared against the dataset to find the most similar match.

5. Similarity Scoring

   • Outputs the matched image, the statue name, and a similarity score.

6. GUI Interface

   • Built with Tkinter:
     • Upload an image
     • Run matching
     • See the result with matched name and score

---

🔧 Tech Stack

• OpenCV
• Pillow
• Tkinter
• JSON