Image Sorter with Gemini AI

This tool uses Google's Gemini AI to analyze images and sort them into different folders based on their potential as stock images and their quality.

Features

• Analyzes images using Gemini AI's vision capabilities
• Sorts images into categories:
  • High/Low stock image potential
  • Good/Bad image quality
• Provides detailed logging of the analysis process
• Handles multiple image formats (PNG, JPG, JPEG, GIF, BMP)

Setup

1. Install the required dependencies:

pip install -r requirements.txt

2. Set up your Gemini API key:

   • Get your API key from Google AI Studio (https://makersuite.google.com/app/apikey)
   • Set it as an environment variable:

     export GEMINI_API_KEY='your-api-key-here'

3. Prepare your images:

   • Create an input_images folder in the same directory as the script
   • Place all images you want to analyze in this folder

Usage

1. Run the script:

python image_sorter.py

2. The script will:
   • Process all images in the input_images folder
   • Create sorted folders under sorted_images/:
     • high_potential/: Images with high stock photo potential
     • low_potential/: Images with low stock photo potential
     • good_quality/: Images with good technical quality
     • bad_quality/: Images with poor technical quality
   • Generate a log file (image_sorter.log) with detailed processing information

Output Structure

.
├── input_images/
│   └── (your images here)
├── sorted_images/
│   ├── high_potential/
│   ├── low_potential/
│   ├── good_quality/
│   └── bad_quality/
└── image_sorter.log

Notes

• Images are copied (not moved) to their respective folders
• Each image can appear in two folders (one for potential and one for quality)
• Check the log file for detailed processing information and any errors