This Python script, objects.py, is a versatile image crawler designed to download a large number of high-quality images from the web based on search queries provided in an Excel sheet. It utilizes the icrawler library, supporting parallel downloading for multiple search queries. The script creates separate folders for each search query, allowing for customization of search parameters to maximize the relevance of downloaded images.

Before running the script, ensure you have the required Python libraries installed. You can install them using pip:

pip install -r requirements.txt

1. Prepare the Excel Sheet: Create an Excel sheet with a column (e.g., 'Objects') containing the search queries. Each row represents a different search query (e.g., objects, personalities, etc.). Save the Excel sheet as 'ListofObjects.xlsx' (or any desired name) in the same directory as the script.

2. Run the Script: Execute the objects.py script to initiate the image download process. The script reads the search queries from the Excel sheet and creates separate folders for each search query in the 'ObjectsImages' directory (you can customize the output directory). It then downloads images corresponding to each search query.

python objects.py

3. Customize Download Parameters: You can customize various parameters within the script to control the download process:

   • initial_num_images_to_download: Set the initial number of images to download for each search query.
   • output_directory: Specify the directory where images will be saved. By default, it's set to 'ObjectsImages'.
   • filters: Adjust filters such as image size, date, color, and type based on your requirements.
   • max_workers: Control the number of concurrent download threads. Adjust this based on your system's capabilities.

By specifying a larger image size or resolution limit in the filters, you increase the chances of obtaining more relevant images. For example:

filters = dict(
    size="large",  # Use 'large' for very large image sizes
    # Other filters...
)

This setting prioritizes larger, higher-quality images which are often more relevant than smaller images.

To expand your search and gather a diverse range of images, you can modify the search_queries list. Add different combinations and variations of your search query:

search_queries = [
    f'{personality}',
    f'people with {personality}',
    f'{personality} in an event',
    # Add more variations or combinations as needed
]

This approach broadens the search scope and brings in a wider variety of images related to your query.

If you want to maximize the number of downloaded images without restricting by size or resolution, simply remove the size filter from the filters dictionary. This will allow the script to download all types of images, including icons, small, medium, and large images:

filters = dict(
    # size="large",  # Remove or comment out this line to download all image sizes
    # Other filters...
)

For more advanced configurations, you can modify the script to use different search engines, adjust the number of images to download, or modify image filters according to your specific requirements. You can interchange the BingImageCrawler and GoogleImageCrawler with each other to get the maximum results.

• The script utilizes the Google Image search engine by default. If you want to use other search engines or sources, you can modify the script accordingly.
• Always be mindful of copyright and licensing when downloading images from the web. Ensure you have the necessary permissions to use the downloaded images for your intended purposes.

This project is licensed under the MIT License - see the LICENSE file for details.

Feel free to customize and extend this script to suit your specific image crawling needs. If you have any questions or need further assistance, please don't hesitate to reach out. Happy image crawling!