TODO

# things todo for openpiv


+ add new dynamic masking to preprocess.py

+ documentation / tutorials / web site
    
    This is always needed to improve. New users 
    should rapidly understand what OpenPiv is capable of. They
    should feel as comfortable as possible on the first `impact`.
    
    A downloads section should be added. This will contain the releases
    example code and example images.
    
    Prepare source distributions for various platform.
    
    Decide where to store OpenPiv-Python website,
    wheter on openpiv.net/python or on Alex github page. 
    Maybe it is better on the former, for gerater flexibility. 
    
    However it seems that the openpiv python documentation
    page is growing fast. Maybe it is better to put clear links
    on openpiv.net to point to three pages for python, matlab, c++.
    
    
+ flow field validation algorithms
    - local median validation
    - local mean validation
    - others?
    
+ iterative image deformation algorithms

+ image-preprocessing, ( filters, masking )

+ hdf5 output

    Writing data to an hdf5 file can be easily done if we do not use 
    multiprocessing. If this is the case, it is much more work and learing.
    
    So basically it would be a first step to: it would be easy for me
    (davide) to create a class in openpiv.tools named something like
    hdf_database_creator which accept a list of files, reads them and save
    data to a single big hdf5 file.
    
+ a module for post processing.
    i have some Object oriented code available at 
    
    http://bazaar.launchpad.net/~lasagnadavide/+junk/PyFlow/view/head:/pyflow/field.py
    
    which already does simple stuff.
    
    Maybe we can start from there.