Fixed a couple of memory leaks in videoInput::getDevice() when multiple devices are present in the system

My build fails with the error:
"C:/PROGRA~1/MINGW-~1/X86_64~1.0-P/mingw64/lib/gcc/x86_64-w64-mingw32/7.2.0/include/c++/cmath:1136:11: error: '::hypot' has not been declared".
I have tried to fix it by adding "#include <cmath>" before the line "#include <Python.h>" but then the build has failed with the error:
"C:/PROGRA~1/MINGW-~1/X86_64~1.0-P/mingw64/lib/gcc/x86_64-w64-mingw32/7.2.0/include/c++/math.h:91:12:: error: 'std::_hypot' has not been declared".
Adding "#include <math.h>" allowed me to build opencv.

Sets a higher limit on videoio's AVI container's maximum chunk size, and adds an explanation of the assertion. Closes #11126

Fix batch norm in training phase.

  * disable carotenete when calling resize on ARM/Aarch64
  * loosen eps on Aarch64

  * Stop calling nppSetStream

* fixed bug #11100 Integer overflow in kmeans

* fixed integer overflow in other divUp-s in kmeans code
fixed warning about size_t to double conversion

* Fix #9363

* Renamed the structure and added a new function to the LineSegmentDetectorImpl class as a static member

* Added a new function to the LineSegmentDetectorImpl class as a static member

color.cpp split (#10869)

* initial split is done

* files renamed (these names are excluded during compilation)

* IPP code moved to corresponding files

* splineBuild, splineInterpolate -> color_lab.cpp

* Lab, Luv: little refactored

* it compiles (didn't check work); Lab OCL code moved to color_lab.cpp

* cvtcolor.cl: Lab/Luv part moved to color_lab.cl

* cvtcolor.cl: color_rgb.cl extracted

* cvtcolor.cl: color_yuv.cl separated

* cvtcolor.cl: color_hsv.cl extracted

* cvtcolor.cl: extracted to color_lab.cl and color_rgb.cl

* helper functions moved to hpp file

* Lab, Luv: moved to color_lab.cpp

* CPU XYZ: to color_lab.cpp

* OCL XYZ: to color_lab.cpp

* warning fixed

* CvtHelper added

* CPU YUV: to color_yuv.cpp, helpers to color.hpp

* CPU HLS/HSV: to color_hsv.cpp

* CPU BGR2BGR: to color_rgb.cpp

* CPU RGB: to color_rgb.cpp

* extra arg removed

* CPU YUV: to color_yuv.cpp

* color code decoded

* OclHelper added, some funcs rewritten

* color_lab.cpp: refactored to use OclHelper

* OCL RGB: to color_rgb.cpp

* OCL HLS/HSV: to color_hsv.cpp

* OCL YUV: to color_yuv.cpp

* OCL YUV planes: to color_yuv.cpp

* OCL: color code reduced

* licence to demosaicing.cpp

* IPP func tables to color_rgb.cpp

* code cleanup

* HAVE_OPENCL ifdefs added

* helpers made more common

* fixed two plane YUV with separate mats

* fixed warning in gcc7.2.0

* precomp header fixed

* color space classification functions fixed

* helpers fixed

* rename: isSRGB -> is_sRGB

* Haiku supporting patches

* Revert uneeded changes

* Whitespace cleanup

This offsets text brightness of pixel brightness values
by offsetting it by 127 to the curent pixel value.
The text is now readable even if pixels are black.

Bug was added in f6ceeaa2 commit, different angle computation functions have different parameter order.

  * add ifdef to apply on specific situation

* Fix #10948

* Add test code

* Fixed build error

* Add value zero

* eigen: test cleanup

Refactor decolor (#11035)

* decolor: modernize deprecated headers (math.h -> cmath)

* refactor contrast_preserve.cpp

* refactor contrast_preserve.hpp (add static/const)

* refactor contrast_preserve.hpp (join assignment and declaration)

* refactor contrast_preserve.hpp (format)

* refactor test_decolor.cpp (indent)

* refactor contrast_preserve.hpp (const)

* contrast_preserve.hpp : optimize by minMaxLoc

* fix trailing-whitespace

* fix warning C4267 on VC++ x64
(conversion from 'size_t' to 'int', possible loss of data)

* use cvRound instead of round_num

* decrease indentation of test_decolor.cpp

* remove pow() to optimize

* remove redundant Mat initialization

Enabling build of visualization module when using VTK 9 (current HEAD of d5bbb9e99bbc6d11d2196c48bfd8f33508554551)

Without proposed change, module throws build error regarding missing `opencv2/cudaimgproc.hpp`

* optimized cv::decolor by removing vector.push_back

* restored CV_INSTRUMENT_REGION() and original indent

* fix build warnings in contrast_preserve.hpp

* undo refactoring contrast_preserve.cpp/hpp

  * avoid misuse of vzeroupper on Visual Studio 2013 Update 5
  * add ifdef to avoid unattended optimization

  * avoid aggressive optimization on Visual Studio 2012 Update 5
  * add code branch to avoid unattended optimization (keep the old code)

* SSE2 : use _mm_cvtpd_epi32 when converting from CV_64F to CV_32S

* No need to define a new universal intrinsic