backported API for storing OpenCV data structures to text string and reading them back (r8481, r8516, r8518, r8522)

fixed some bugs in gpu module under Windows:
* convertTo alignment error
........

Bug #1991 : fixed gpu interpolation (remap, resize, warp...)
........

fixed several problems with CUDA 5.0
* gpu::LUT, uses device memory instead of host memory
* gpu::multiply, round mod for CV_8U depth
........

compilation with CUDA5.0 (#1982, thanks to Rodrigo)
........

fixed possible compilation error for macos
........

Added check for CUDA toolkit version ( older can't compile Kepler )
........

get back to single-version ffmpeg wrapper (no more v2); make sure the library builds and the tests pass on Ubuntu 10.04 that includes pretty old version of ffmpeg

get back to single-version ffmpeg wrapper (no more v2); make sure the library builds and the tests pass on Ubuntu 10.04 that includes pretty old version of ffmpeg

avoid simultanous reading data from 2 videos Highgui_Video.regression, since earlier versions of ffmpeg (or the way our wrapper uses them) can not handle that properly

force BUILD_PNG/JPEG/TIFF/ZLIB on MacOSX, because sometimes the found lib<...> is different from the one used in build.