I can "describe" this race track precisely by imposing an X Y grid on top of it with enough resolution to capture each of your little dashes. Then report the coordinates and colors of each occupied point in the grid.
Effectively, that is what you have done with your "image", the computer screen is just a square XY grid of pixels,which can each be some color. That is how television and computer videos are made, they are not continuous but rapidly displayed still frames that create the illusion of movement.
The race track can be compressed to ignore the squares on the grid that contain no element of race track, and the squares of the grid can be relatively large, compared to pixels. Say, 1/4 of the width of the track.
An alternative would be to recite exact measurements, length, width, and angles and curvatures, but there would be many of those required as well to produce a copy a jury agrees is the same image as the original.
Otherwise, if your "description" prohibits reciting a long accurate accounting of the image, then of course it is unlikely it will be accurately reproduced.