This program uses cellular automata to grow two dimensional, semi-random, tree like structures. The demo shows how it can be used for landscape generation in an interactive scenario.
The node.js interactive demo runs in your terminal. The tree generation algorithm is provided in both node.js and python.
Clone and run.
git clone https://bitbucket.org/liamilan/forest-generation-cellular-automata.git
npm startThis will run an interactive demo where you drive your car in a savanna of cellular automata generated trees.
The cellular automata's cells have one of two states, living, or dead. Each cell can grow up, left, and right in each turn according to the following probabilities.
- The probability to grow left is h / 30 out of 1, with h being the height of the cell starting at 0.
- The probability to grow right is also h / 30 out of 1.
A cell is more likely to grow outward the higher it is. In real trees, this helps outgrow other trees, making sure the leaves can catch sunlight.
- The probability to grow up is (1 - (h / 30)) * 0.125 out of 1.
- A cell will not grow if there is a cell alive somewhere above it (or if it is not alive).
It is vital for real trees to catch sunlight to grow.
To run the tree generation node.js version of the automata:
node automata/automata.js [number of cells to start with]
To run the tree generation python version of the automata:
python automata/automata.py [number of cells to start with]
I am Liam Ilan, a 12 year old software developer who is never working, but always playing around.