The idea of superpixels was originally developed by Xiaofeng Ren and Jitendra Malik [1]. This implementation is different, and is a version of that used in [2],[3]. See the README for more information.
Update (March 11, 2010): 64-bit modifications of code available (thanks to Richard Lowe for providing the fixes)
Update (March 7, 2006): Fine scale superpixel code [3] now available.
Tarball and directory of 64-bit superpixel code.
Tarball and directory of superpixel code.
Example images (from [3], N_sp=1000):
|
|
|
|
|
|
|
|
|
|
|
|
[1] X. Ren and J. Malik. Learning a classification model for segmentation. In Proc. 9th Int. Conf. Computer Vision, volume 1, pages 10-17, 2003.
[2] G. Mori, X. Ren, A. Efros, and J. Malik, Recovering Human Body Configurations: Combining Segmentation and Recognition, IEEE Computer Vision and Pattern Recognition, 2004.
[3] G. Mori, Guiding Model Search Using Segmentation, IEEE International Conference on Computer Vision, 2005.