Generative Image Composition is realized by the algorithm written in Processing. The idea is to create certain unpredictability and uniqueness into the photo canvas. The computer program follows the rules determined by us.
The rules are:
- images should compose a continues path, and thus, compose a form
- choose between 3 sizes of an image
- finish composition when out of canvas or no possibility to place the next image
We have a number of next ideas. First, we were thinking of human intervention to the algorithmic composition that still follows the rules of the algorithm. Even more, we would like to turn it into a kind of collage act and social game.
Here is the documentation of manual generative composition:
Instruction for 2-player-game:
tools: dice, 3 colors of paint, 3 sizes of squares’ templates (corresponding to the size of images)
1. Throw a dice that will determine the size of square to be painted (1&2-> S-size, 2&4-> M-size, 3&6-> L-size)
2. Paint the chance-based square on the canvas that it touches the edge of image and/or the square of the same color. To touch the square of an opponent is not allowed.
3. The game is continued and generative canvas filled with painted squares until there are no possibility to paint determined size square.
4. The winner is the one who painted the last square on the canvas.
Compositions that combine manual and mechanical technique of generative composition following the same algorithm.
We have presented the 3 different executions in our solo exhibition called Execution#X in IAMAS OS, Ogaki, Japan.
Another 3 executions of photo canvases were completed through participative performance. In the opening we asked audience to play the game against each other, and thus, to generate manually the compositions.
Another 3 executions with a different canvas size:
Talking about the next step, we are going to make a generative image composition from photos on social networks!