Poobesh Gowtham an Industrial Design Undergrad at the IDC School of Design, IIT in Bombay created an awesome color picker project, Vangogh. The website describes Vangogh as a “search engine for color palettes”. If you want to find a palette that embodies an idea, theme or word you can use Vangogh to suggest several palettes. The image below shows what we get when we search for ‘Adafruit’.

Once a query is submitted to the site, it searches Bing for 100 relevant images. These images are clustered based on their color similarity using k-means. Once similarly colored images are grouped together, they are grouped into a collection of colored pixels and mapped in 3D space using R,G,B as X,Y,Z. Then, k-means is used a second time to predict color clusters. Each group of 100 images is used to generate 5 themes of which you can choose 3 to 7 colors for each theme.

Each image is then 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 colored points is then grouped using K-Means clustering, to generate the color palette. Palettes are tweaked and ranked based on color theory.

Just to compare, the picture above shows 5 basic color themes and the picture below shows some of the top results for ‘Adafruit’ image search on Bing. Looks pretty accurate although I would love to see a shock of pink in there too :). If you’d like to learn more about how Vangogh was created take a look at the ‘how it’s made’ page or checkout Poobesh Gowtham on GitHub.

Written by Rebecca Minich, Product Analyst, Data Science at Google. Opinions expressed are solely my own and do not express the views or opinions of my employer.