Removing all shadowing variables, make functions to const, make the table of color-names become static

Fixed: ROI extraction when the given ROI is out of image; made the max_patch_size to be adjustable; add the CN feature extraction method

+ add KCF Tracker, initial commit, added: tutorial, trackerKCF.cpp, modified: tracker.cpp, tracker.hpp

Replaced property macros in headers

Bug fix #1

Removed sphinx documentation

The bug appears on the initialization step, if the initial bounding box width was larger than height. The problem was in function calculating a binary code for ensemble classifiers. In this function “measurement” values were used in wrong order (h1,w1,h2,w2; instead of: w1,w2,h1,h2), generating a access violation error on “patch” data access by the reason that width>height and pointer to data is calculated linearly data.step*height+width

fix missing delete[] calls in or_mnist.cpp

Java and python wrappers for contrib modules

Function cuda::integral() does not require an extra buffer any more.

Remove algorithm factory