major time consuming part for training app is in function fillPassedSamples for negatives, 
this change make SetImage quicker, in a test of my own data, the total time for SetImage decrease from 9177666 to 5839263,
only help for Haar feature and non Haar::ALL mode which is the most commonly case

minor fix since predict() returun int

make the software system dependent
add LICENSE on top

opening a folder is system specific - made system specific code
added license

2. Algorithm::load/save added (moved from StatModel)
3. copyrights updated; added copyright/licensing info for ffmpeg
4. some warnings from Xcode 6.x are fixed

work_var_count and sample_count are both 32bit integer, if the product of work_var_count and sample_count is bigger than 2^31, the compiler will treat (work_var_count*sample_count) as a negative number. Force work_var_count as an unsigned 64bit integer to avoid this issue.

Check sure that we're not already below required leaf false alarm rate before continuing to get negative samples.

Documentation improvement
Bug in output format for JPG set fixed