Is this a color detection API or a palette extractor?

The API performs color detection from image and returns a clean palette of dominant colors with HEX/RGB values and their share (%) in the image.

What does min_share mean in color detection?

min_share defines the minimum percentage of the image area a color must cover to be included in the result. Colors below this threshold are filtered out to remove noise and parasite pixels.

How does the API avoid noisy or parasite pixels?

The algorithm groups visually similar pixels into clusters, calculates their coverage, and returns only dominant color clusters above the configured min_share threshold.

Can I control the number of returned colors?

Yes. You can limit the output using the max_colors parameter and filter small color clusters using min_share.

Do you support transparent PNG images?

Yes. Transparent areas are ignored by default and do not affect the extracted color palette.

Color Detector API: Detect Color From Image and Extract a Clean Color Palette

Upload an image. Get back a list of dominant colors with HEX/RGB and their share (%) in the image — with noisy/parasite pixels filtered out. This is an image color API built for reliable color detection from image in production workflows.

Request free demo access to test the API on your images

What you get

Image color recognition focused on dominant colors (not random single pixels)
Extract color palette from image as HEX + RGB
Color share (%) for each returned color (coverage in the image)
Configurable minimum share threshold (e.g., return only colors > 5%)
Stable JSON output, ready for automation and onboarding flows

Example request

POST /v1/color-detector
Content-Type: multipart/form-data

image=@photo.jpg
min_share=5
max_colors=10

Example response

{
  "min_share": 5,
  "max_colors": 10,
  "colors": [
    { "hex": "#1F2937", "rgb": [31, 41, 55],  "share": 42.8 },
    { "hex": "#F9FAFB", "rgb": [249, 250, 251], "share": 31.5 },
    { "hex": "#D97706", "rgb": [217, 119, 6], "share": 12.2 },
    { "hex": "#111827", "rgb": [17, 24, 39], "share": 8.1 }
  ]
}

How the algorithm avoids “parasite pixels”

Many “detect image color” tools overreact to tiny details: compression artifacts, reflections, sensor noise, or 1–2-pixel speckles.

This color detector API focuses on dominant colors by:

Grouping visually similar pixels into stable color clusters
Computing coverage for each cluster and returning color share (%)
Filtering out clusters below min_share (e.g., ignore everything under 5%)

Result: a cleaner palette for design systems, product onboarding, brand color extraction, and UI automation.

Use cases

Brand onboarding: detect color from image (logo/screenshot) and auto-fill theme colors
Design tooling: extract color palette from image for asset pipelines
E-commerce: derive dominant colors for filters (“red”, “blue”, “neutral”) and tags
Content automation: image color recognition for templates, thumbnails, and banners

Pricing

Volume-based
Starting from €9/month
Request free demo access to evaluate the quality of the results

Mini story

We first built this algorithm as an internal production component for product onboarding — to automatically detect brand colors from uploaded images. After running it reliably in real workflows, we made it available as a standalone Color Detector API.

FAQ