cuda: move CUDA modules to opencv_contrib: https://github.com/opencv/opencv_contrib/pull/1781

OpenCV 4.0+

* core: drop usage of CV_USRTYPE1 in OpenCV

avoid OpenCV crashes due size change CV_ELEM_SIZE(CV_USRTYPE1): 8 -> 2

* ! fix persistence internal types

Before this fix, the code would fail if only standard deviations of
extrinsic parameters are requested. While standard deviations matrix
should be computed if any set of standard deviations is requested. A
variable is added to represent this case.

- DNN: avoid very large models
- build filter per module
- fixed longTestFilter

Updating the base SDK from 6.0 to 8.0 where applicable

- ffmpeg 4.0.2
- openh264 1.8.0

Support asymmetric padding in pooling layer (#12519)

* Add Inception_V1 support in ONNX

* Add asymmetric padding in OpenCL and Inference engine

* Refactoring

* feat: move gcgraph to detail

* chore: puts gcgraph cv::detail namespace

* matlab: move to opencv_contrib

* cmake: preserve variables scope for processing modules

- use macro instead of function to avoid scope resets