Experimentation
Our first goal was to ensure that our snake was working properly. The snake appears to be very sensitive to initial parameters, so for the length of the experiment, we are optimizing inputs for all algorithms (Shah, GVF, threshold, and region grow).
We decided on a variety of test images to check what type of features snakes are good at, and what types they are bad at. Three are shown below:
 |
 |
 |
 |
 |
 |
By examining the following comparison chart, we can draw our experimental conclusions: