Project 1: Segmentation

Implement and report results over at least two of the three approaches to Segmentation.
1. MRF (Chellappa et. al, Dudes et. al., Manjunath et. al, Derin and Elliot)
2. Graph Cut (Shi Malik)
3. Filtering (Malik Perona or Wavelet based methods)

Submission Deadline :- Nov. 8

Report describing in detail training (if supervised), choice of features, results and comparison with ground truth (when available).

Dataset:

Please cite the datasets accordingly.
1) MicroTextures
http://www-dbv.informatik.uni-bonn.de/image/segmentation
Training:
http://www-dbv.cs.uni-bonn.de/image/texture.tar.gz
Testing:
http://www-dbv.cs.uni-bonn.de/image/mixture.tar.gz
2) For Graph Cuts: (Use Color/Intensity Features)
Berekely Segmentation Dataset:
http://www.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/
3) Julesz Textures: (to be updated)
http://www.citr.auckland.ac.nz/~georgy/research/texture/thesis-html/img9.png

Figures from Perona-Malliks's Paper Please email me (Aswin) if this doesn't work out in your experiments.