Snakes: Active contour models

Snakes: Active contour models is a highly cited research paper in Computer Vision which introduced an algorithm to detect edges in images using gradients. Published in 1988, the abstract explain that snakes are "energy-minimizing splines guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges." In the picture on the right, the user has drawn a snake and has applied three springs or hard constraints to the spring.

The Project

I was on a team with three other graduate students at ASU tasked with implementing the snakes from the paper. We implemented hard and soft constraints (also called springs) into our final program, which allowed the user to interact with the snakes in real time. The interactivity gives the power to pull a snake out of a false local minima and find the desired edge.

Results

The team successfully implemented the snake algorithm in MATLAB with full hard and soft constraint functionality. Take a look at the video demo of the software we built and the presentation about the snake algorithm.

• Snakes Presentation

What I Learned

1. How to read and write research papers
2. Image Processing in MATLAB
3. A strengthened love for Computer Vision