put the detection-based tracker back (to objdetect module), so that the Android face detection sample builds fine. The patch by Konstantin Matskevich that adds C++11 threads support has been merged in too. Also, fixed compile error in the doc builder.

removed contrib, legacy and softcsscade modules; removed latentsvm and datamatrix detector from objdetect. removed haartraining and sft apps.
some of the stuff will be moved to opencv_contrib module.
in order to make this PR pass buildbot, please, comment off opencv_legacy, opencv_contrib and opencv_softcascade test runs.

reshape input matrix, since the function works with data
as with [N x dims] matrix

Add MatlabFormatter for matlab style output.
See http://code.opencv.org/issues/2789

Python typdef fixed. Reverted the example to MOG2. Not time to make the command line switch for now.

literalinclude

literalinclude, dropped :lines:

Env setup for testing package implemented using /etc/profile.d;
Variable with path for all native samples added;
Path for test binaries and test data updated.

get_hogdescriptor_visu is passed img->clone(), which is not usable
as a non-const reference (and it doesn't modify the argument, anyway).

Fixed a valgrind 'Conditional jump or move depends on uninitialised value(s)' on cv::kmeans(...). The original code used points(sampleCount, 1, CV_32FC2), which confused generateCentersPP into thinking it is a 1 dimensional center, instead of 2. As a result it would set only the x variable and leave y unitialised.