Color Me Pythonic: Extracting Dominant Colors from Images with PIL and K-Means

What color is that shirt, really? In this talk, we'll dig into how to extract
dominant colors from clothing images using Pillow and scikit-learn's K-Means
clustering — drawn from a real browser-based wardrobe app built as a college
research project.

You'll walk away knowing how to:

• Resize and preprocess images for efficient clustering
• Run K-Means on pixel data to identify a dominant color palette
• Handle tricky edge cases like busy patterns and low-contrast photos
• Surface color results in a usable, visual format

No prior computer vision experience needed — just Python curiosity and maybe a
messy closet.

Presented by

Veronica k McNeece