Duplicate Image Deleter

This project is a Python-based application that identifies and removes duplicate images between two directories (e.g., Google Drive and OneDrive). The program uses a graphical interface to allow users to manually confirm deletion decisions or skip them.

Features

Enhanced Features:

1. Multi-Year and Multi-Month Directory Support:

   • Handles directories for multiple years and months dynamically, ensuring a comprehensive comparison process.

2. Preprocessing for Performance:

   • Precomputes image pairs and caches resized grayscale versions for faster SSIM calculations.

3. Flexible Threshold:

   • Configurable similarity threshold for SSIM-based image comparison, enabling sensitivity adjustments.

4. Modernized GUI:

   • Displays both images flagged as duplicates side-by-side with options to keep, delete, or cancel.
   • Enhanced user interface with progress bars and dynamic updates.

5. Threaded Preprocessing:

   • Uses multithreading for preprocessing image pairs, ensuring the main application remains responsive during calculations.

6. Improved Logging and Error Handling:

   • Logs detailed error messages during image loading, resizing, or SSIM calculation.

7. Configurable Deletion Behavior:

   • Allows users to select whether to delete duplicates from Google Drive or OneDrive.
   • Option to enable or disable confirmation prompts for deletion actions.

---

Requirements

• Python 3.8 or later
• Libraries:
  • os
  • tkinter
  • cv2 (OpenCV)
  • Pillow
  • skimage
  • shutil
  • random

Install required packages:

pip install pillow opencv-python scikit-image

---

How It Works

Step 1: Setup Directories

The CreateDirs script creates year/month subdirectories under ./Google and ./OneDrive and populates them with random images from a TestPhotos directory.

Step 2: Compare Images

• The program iterates through the specified year and month directories.
• For each pair of images, SSIM is used to calculate their similarity.
• If the similarity exceeds a threshold (default: 0.95), the images are flagged as duplicates.

Step 3: User Interaction

• The user is presented with both images in a Tkinter window.
• Options include:
  • Keep one of the images.
  • Delete one of the images.
  • Cancel the deletion process.

---

Usage

1. Setup Your Files:

   • Place the TestPhotos folder with sample images in the root directory.
   • The script will populate ./Google and ./OneDrive directories with random subsets of these images.

2. Run the Program:

   • Start the script by running:

     python DuplicateDelete.py

   • The GUI will appear for flagged duplicate images, allowing you to decide on each case.

3. Configuration:

   • Modify the following variables in the script to suit your needs:
     • delete_from_google: Set to True to delete images from Google, False to delete from OneDrive.
     • confirm_delete: Set to True to enable confirmation prompts before deletion.

---

File Structure

.
├── CreateDirs.py          # Script to create and populate directories
├── DuplicateDelete.py     # Main script for detecting and deleting duplicate images
├── Google/                # Google Drive simulation folder
├── OneDrive/              # OneDrive simulation folder
├── TestPhotos/            # Source folder containing sample images
├── README.md              # Project documentation (this file)

---

Example Output

1. Folder Structure: After running CreateDirs.py, the directories are organized as:

   Google/
   ├── 2025/
   │   ├── Jan/
   │   ├── Feb/
   │   └── ...
   ├── 2024/
   └── ...
   OneDrive/
   ├── 2025/
   │   ├── Jan/
   │   ├── Feb/
   │   └── ...
   ├── 2024/
   └── ...
   

2. GUI Window:

   • Displays both images flagged as duplicates.
   • Options to keep, delete, or cancel.

3. Progress Bar:

   • Real-time progress updates during image preprocessing and SSIM calculations.

---

Notes

• Performance:
  • Comparing large image sets may take time. Optimize by reducing the number of images or directories.
• Error Handling:
  • Provides detailed error logs for missing files or calculation issues.

---

Feel free to modify and adapt the code to your specific use case!