Scissors

Intellegent Scissors

Interactive tool for image segmentation

Contents

Introduction

Introduction

This Intelligent Scissors were written as a project in the Intelligent Systems Lab under the supervision of Prof. Gershon Elber

The problem

Digital image segmentation techniques are used to extract image components from their sorrounding natural background and identifying their boundary.

Current solutions for this problem:

Automatic image segmentation tools - these tools are currently researched, and as far as I know they require many assertions in order to produce useful results.
Snakes - this method which is based on Osher and Sethian level set method require the user to specify the assumed boundary and then it improves the boundary. Unfortunately the method uses two phases: get boundary from user.and improve the boundary, which would take some time and if the user received unacceptable result he/she needs to go over the process again.

The aim of this project was to implement an intelligent scissors tool, based on a paper of Eric N. Mortensen and William A. Barrett from Brigham Young University. (see Bibliography). The fundamental technique of this method is to convert the problem to a graph search problem.

Downloads

Source Code

Windows NT

Executable

Windows NT

Bibliography

Eric N Mortensen, and William A. Barrett "Intellegent Scissorfro Image Composition" SIGGRAPH 1995, 191-198.
M Kass, A Witkin, and D Terzopolous. "Snakes: Active contour models". International Journal of Computer Vision, 1:321-331, 1988
L D Cohen and R Kimmel. "Global Minimum for active contours models: A minimal path approach" International Journal of Computer Vision", 24(1):57-78, 1997

Author

If you have any questions, comments or bug reports please contact Ronen Shor