Lets the user choose the maximum number of iterations the robust
estimator runs for, similary to findHomography. This can significantly
improve performance (at a computational cost).

* Custom layers with nGraph

* nGraph: multiple outputs from nodes

* Fixed getCPUCount

Minor new line changes

Android fix | efficient linux checks

Android fix 2

Fixed cpu logic for non linux platforms

Android fix 3

Android fix 4

* No v1 case handle | Refactor long lines

* Refined Cgroups logic | Combine Android and Linux

* Fixed directives

* Added support for --cpus | Fixed minor bug in Andriod | Change file read logic

* Added macro checks for apple errors

* Fixed macro to include android

* Addressed review comments

* Fixed android macro

* Refined return values

* Fixed apple warning

* Addressed review comments

* Fixed whitespace

* Android Fix try 1

* Android Fix try 2

* Android Fix try 3

* Removed unwanted endif

* Android Fix try 4

* Android Fix try 5

* Macro Restructure

* core: updates to CPUs detection (minor)

* Update to new ICV packages

MacOS ia32 ICV package is no longer built.

* cmake(ippicv): add Apple 32-bit check

* Updated commit hash

* add  python version of video-input-psnr-ssim

* remove ret

* documentation changes

* added link for python file

* command line argument

* dnn: fix processing of input shapes

- importer: avoid using of .setInput() => .setInputShape()
- setInput: shape limitation check (partial)

* dnn(test): test .setInput() in readNet()

fixed the ordering of contour convex hull points

* partially fixed the issue #4539

* fixed warnings and test failures

* fixed integer overflow (issue #14521)

* added comment to force buildbot to re-run

* extended the test for the issue 4539. Check the expected behaviour on the original contour as well

* added comment; fixed typo, renamed another variable for a little better clarity

* added yet another part to the test for issue #4539, where we run convexHull and convexityDetects on the original contour, without any manipulations. the rest of the test stays the same

* fixed several problems when running tests on Mac:
* OCL_pyrUp
* OCL_flip
* some basic UMat tests
* histogram badarg test (out of range access)

* retained the storepix fix in ocl_flip only for 16U/16S datatype, where the OpenCL compiler on Mac generates incorrect code

* moved deletion of ACCESS_FAST flag to non-SVM branch (where SVM is shared virtual memory (in OpenCL 2.x), not support vector machine)

* force OpenCL to use read/write for GPU<=>CPU memory transfers on machines with discrete video only on Macs. On Windows/Linux the drivers are seemingly smart enough to implement map/unmap properly (and maybe more efficiently than explicit read/write)

Adding python codes to samples/python

* Python version of samples:
- laplace.py
- drawing.py

* Update drawing.py

* Update drawing.py

* Update drawing.py

* Update laplace.py

* Update laplace.py

* Update drawing.py

* Update drawing.py

* Update laplace.py

* samples

* drawing

* drawing

* drawing.py