This is a quick code experiment that visualizes images as trajectories in color space, one line per pixel row.
RGB space is the cube formed by all the possible value combinations of the red, green and blue axis.
HSB is the cube formed by hue, saturation and brightness.
When the pixel rows are drawn as trajectories in one of these spaces, contrasting colors among adjacent pixels will result in longer jumps within the space, and will therefore become much more visible.
Thus, this visualization method brings forward color variations, but also the general shape of the color palette the image is using.
Lines are tinted according to their departure location.
You can see more images at the Flickr set, or download the source code and user your own images.
To run the sketch, you will need Processing, which you can download here, and the excellent PeasyCam library by Jonathan Feinberg
Sorry, no online version because OpenGL makes it difficult. Happy 3D walks!
