readme

phow is the code for extracting pyramid features from image

Pyramid Histogram of visual Words - PHOW
=======================================


Introduction
~~~~~~~~~~~~

This package contains software for computing the pyramid histogram of visual words using gray level or color information.
If you use this software for your research, please cite [1] as a reference.
We refer to [1-3] for a brief explanation of how the pyramid works,
and to [2-3] for a detailed explanation of its extension over colour images.


Compatibility
~~~~~~~~~~~~~
The code requires Matlab v 7.0 or later to run.
you might need to download the executable file to compute the SIFT features:

http://www.robots.ox.ac.uk/~vgg/research/affine/descriptors.html#binaries



Quick start
~~~~~~~~~~~

Let <dir> be the directory where the package was uncompressed.
Start matlab and enter in the matlab shell:

1) cd <dir>
2) main

The main program will compute the phow descriptor and save it as a "txt" file.


Remark
~~~~~~

You can modify the code to compute the descriptor over a roi. Open the do_phowDescriptor.m file and change

roi = [1;size(I,1);1;size(I,2)];

by the roi position in the image.


Support
~~~~~~~

For any query/suggestion/complaint or simply to say you like/use this software, just drop us an email

[e-mail]

References
~~~~~~~~~~

[1] A. Bosch, A. Zisserman and X. Munoz.
Scene Classification Using a Hybrid Generative/Discriminative Approach
IEEE Transactions on Pattern Analysis and Machine Intelligence (2008) 

[2] A. Bosch, A. Zisserman and X. Munoz.
Scene Classification via pLSA
Proceedings of the European Conference on Computer Vision (2006) 

[3] S. Lazebnik, C. Schmid, and J. Ponce.
Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories.
Proc. IEEE CS Conf. Computer Vision and Pattern Recognition (2006)