When you search for a color palette, VANGOGH queries the Brave Search API and downloads 100 images tagged with the search term. It then resizes the images to 50×50 pixels to improve efficiency. The images are cached and can be viewed here.

The images are clustered based on perceptual similarity, i.e., similar colored images are grouped together. This is done by generating color moments for each image, which serve as signatures and are then used to cluster using K-Means.

Each image is converted to a collection of pixels. The constituting colors are visualised on a 3D space, where the spatial coordinates X, Y, Z are the values R, G, B of the pixel. The collection of coloured points is grouped using K-Means clustering to generate palettes. Palettes are tweaked and ranked based on colour theory.

Each image now has a palette. Palettes from the same image cluster create a palette theme as they share a common colour scheme. VANGOGH provides four themes with five palettes in each. The palette data and the 3D visualisation of the colour space of the images are sent to the browser in just a few seconds.