A blog about algorithmic art and fractal aesthetic. Click here to subscribe to the RSS feed .
December 6th 2009
Charis Tsevis
Charis Tsevis creates portraits from a mozaic of smaller images. He's not the only artist using this technique, but I find his works have something special. The idea to create such an image seems to be the following. Start from a pool of square images, sorted by their average color (or simply their average brightness in the case of a black and white image). Then decompose the portrait into squares ("pixels", but not necessarily of the same size), compute the average color in each square and replace it by the image in the pool having the closest average color.
What makes Tsevis's works interesting is that the uniformly colored regions are covered by large pixels, while small pixels are used in regions requiring details. The large pixels give rise to large copies of the small images, which interfere nicely with the portrait. Tsevis also offer large resolution pictures on his Flickr account, and I converted two of them into zoomable images using Seadragon. Here they are: