This should be safe todo unless we are writing a signal handler.

xioctl() assumes that ioctl takes int request. Cope with
  int ioctl(int d, unsigned long request, ...)
to avoid "invalid argument".

* Decouple Video4Linux2 support from Video4Linux as existence of
  v4l2 on a system does not imply support for v4l.
* Don't use V4L's struct video_window in V4L2 code.
* Removed __USE_GNU as comment says:
      /* support for MJPEG is only available with libjpeg and gcc,
         because it's use libjepg and fmemopen()
  so replace with test for fmemopen() if found necessary.

The video(4) driver provides a Video4Linux2 compatible API for
various video peripherals. This patch propagates HAVE_VIDEOIO if
the sys/videoio.h include file is found, which is the signature of
video(4).

Fixed visualization by choosing the color appropriate to the detection
score.
Previously the example showed all detections with the same color
disregarding the confidence. This led to the impression that the object
detection did not work at all because there are many detections with low
confidences.

PR to master was
https://github.com/Itseez/opencv/pull/320

Introduced new Java API
void org.opencv.core.Core.setErrorVerbosity(boolean verbose)
used to suppress output to stderr from OpenCV's asserts

Common part of all source files is extracted to special header

- add syntetic images generation function to ts
- use generated syntetic images