Texture Similarity App

A Python-based desktop application that compares two images and computes their texture similarity using classical texture descriptors: LBP, GLCM, Gabor filters, and Entropy.
The final similarity score is calculated using Euclidean distance and displayed in a simple, user-friendly interface.

🧠 Features

• 📷 Select and compare any two images
• 🔍 Uses four powerful descriptors:
  • LBP (Local Binary Pattern)
  • GLCM (Gray-Level Co-occurrence Matrix)
  • Gabor filters
  • Entropy
• 📊 Computes a unified similarity score (%)
• 💾 Automatically logs comparisons in a .txt file
• 🖼 Visual side-by-side comparison with Matplotlib
• 🪟 Simple GUI built with tkinter

📄 Documentation

For a full explanation of the algorithms, implementation details, and theoretical background, see the full project report:

📘 Documentation (PDF)

🛠 Technologies Used

• Python 3.7+
• OpenCV (cv2)
• NumPy
• Scikit-Image
• SciPy
• Matplotlib
• Tkinter (for GUI)

🚀 Installation

1. Clone the repository:

git clone https://github.com/svuksanova/dpnsProject.git
cd dpnsProject

2. Install dependencies:

pip install opencv-python numpy scikit-image matplotlib scipy

3. Run the main script:

python main.py