Playing with Snakes

As part of my Introduction to Computer Vision class, I got to implement and analyze an algorithm of my choice. I ended up picking the Greedy Snake algorithm as described by Williams and Shah. Snakes are active contour models which, when positioned near an object, tend to converge to its edges. They are thus useful to track or find objects within an image given a decent initial guess. One interesting application of this algorithm is in video tracking, where the output from one frame can be used as the input to the following frame, making it possible to track the object’s position from frame to frame cheaply.

If you are interested in the details, I posted the MATLAB code as well as the analysis on Github, so go ahead and fork it!