Extract dominant colors and full palette from any image
This tool uses Canvas pixel analysis with color quantization. The image is drawn onto a hidden canvas scaled to 200px max dimension (for performance), then every pixel's RGB values are quantized to 32-step buckets (reducing 16.7M possible colors to a manageable set). A frequency map counts occurrences of each quantized color, sorted by dominance. The pixel picker operates on the full-resolution canvas using getImageData(x, y, 1, 1) for precise single-pixel sampling with correct DPI scaling calculations.
Real-world use cases:
This tool is part of the FAK LAB ecosystem, founded by Faizan Ahmad Khan Khichi. Images are processed 100% in your browser using Canvas API pixel manipulation. Your images — including proprietary designs, unreleased artwork, or confidential materials — are never uploaded to any server. The pixel data exists only in your browser's memory and is discarded when you navigate away.
Color quantization rounds RGB values to 32-step intervals for efficient grouping (e.g., rgb(129,65,33) becomes rgb(128,64,32)). This groups similar shades into clusters rather than listing millions of unique pixel values. The pixel picker gives exact, unquantized colors for precision when you need the true value at a specific point.
The tool displays up to 20 dominant colors sorted by pixel frequency. Most images have 5-15 visually distinct dominant colors. The quantization algorithm ensures results represent meaningful color regions rather than noise or single-pixel variations.
Yes. Pixels with alpha values below 128 (semi-transparent to fully transparent) are excluded from the dominant color analysis to prevent transparent areas from skewing results. The pixel picker still shows the alpha channel value for any sampled point, so you can identify transparency levels.