A small python package for detecting NSFW content on a Windows display.
This package does the following by default:
- Take a screenshot on all connected displays. You can take a screenshot of just the active window by passing
input='active_window'
to thecheck
method. You can also pass in a Pillow image like so :input=Image
- Detect actual images from the screenshot. You can skip this by passing
parse_images=False
to thecheck
method. - Run skin detection. You can set the
skin_threshold
parameter to have NSFW checking skipped if an image contains very little skin. - Use AI to detect how NSFW each image is. You can pass
check_type='skin'
to just get a skin color score. - Returns each analyzed picture along with its results.
# Import and initialize the main Checker class
from nsfw_screen_detect import Checker
checker = Checker()
# Run check on all monitors
results = checker.check()
# Run check on active window
results = checker.check(input='active_window')
# Run check for skin content
results = checker.check(check_type='skin')
# Run check without parsing real pictures from screenshot
results = checker.check(parse_images=False)
Big thanks to NudeNet for their AI model.