中文

Amazon Dogs - Dog Image Downloader

A simple open-source tool for downloading dog images from Amazon's 404 error pages.

What is this project?

This project downloads dog images that appear on Amazon's error pages. These images are accessible via predictable URLs like:

• https://images-na.ssl-images-amazon.com/images/G/01/error/1._TTD_.jpg
• https://images-na.ssl-images-amazon.com/images/G/01/error/2._TTD_.jpg
• And so on...

The project includes 200 pre-downloaded dog images ready to use.

How to use the downloaded images?

Option 1: Clone the entire repository

git clone https://github.com/good-sellers/amazon-dogs.git
cd amazon-dogs

The downloaded images are located in data/cleaned_dogs/images/:

• 200 dog images (cleaned_dog_1.jpg to cleaned_dog_200.jpg)
• Index file: data/cleaned_dogs/cleaned_index.json

Option 2: Download images only

You can download just the images from the repository without cloning the entire codebase.

How does it work?

The downloader works by:

1. URL Pattern: Amazon uses a predictable URL pattern for error page images

   • Base URL: https://images-na.ssl-images-amazon.com/images/G/01/error/
   • File format: {number}._TTD_.jpg (starting from 1)

2. Sequential Fetching: The crawler tries image URLs in sequence (1, 2, 3, 4...)

3. Smart Stopping: Stops when:

   • Reaches max number of images (default: 1000)
   • Gets 100 consecutive 404 errors
   • 3-second delay between requests to avoid rate limiting

4. Image Storage: Successfully downloaded images are saved to data/dogs/ with an index file

Optional: Run the crawler yourself

If you want to download more images:

# Install dependencies
pip install -r requirements.txt

# Run the crawler
python dog_crawler.py

License

MIT License