small improvement to landmarks_demo.py

the landmarks_demo.py fails with an obscure error when no faces are detected in the input image. Added some error handling so that it's more transparent.

Added Java bindings for dnn_superres module

fix: remove using namespace cuda from public header

Fixed several test failures in Python tests for CUDA modules

`cudacodec::VideoReader` return luma hist from `nextFrame` if requested

Add Ubuntu 22.04 to CI.

Wstringop-overflow warning fix in bio inspired module

Fixed memory leak in EllipseDetector and Mat addressing

Modified histogram kernels to work with non aligned data

FFmpeg/4.x: update FFmpeg wrapper 2023.6

`cudacodec::VideoReader` - fix CUDART_VERSION tests

`cudacodec::VideoReader` FORCC update with main #23540

`cudacodec::VideoReader` - ensure `FormatInfo.valid == true` on return from constructor

Moved barcode module to opencv/objdetect

`cudacodec::VideoReader` add reconfigure decoder functionality

Add missed parameters descriptions

---------
Co-authored-by: NMaksim Milaschenko <m.milaschenko@yadro.com>
Co-authored-by: NMaksim Shabunin <maksim.shabunin@gmail.com>

moved aruco_dict_utils.cpp from opencv_contrib to opencv

Promote cv::VideoCapture properties in VideoReader

Wechat qrcode perfomance tests