Bright edges source code and demo (#1449)

* Bright edges source code and demo

* Fix mac errors

* Upload again because of null characters

* Convert to UTF-8

* Change cv::WindowFlags::WINDOW_AUTOSIZE

* Remove parameter

* Remove trailing white spaces (documentation builld error

* Remaining white spaces

* Correct warning

* Correct surround

* Add comment

* Moving prototype out of header

* Add ximgproc namespace, fix example.

* Fix namespace for example

* Remove windows desctruction

* Correct Mat declaration

* Removed unused gray

* Change local functions to static

* Remove tabs and replace by 4 spaces

* Add comments

* Fix comments

* Retry for random test stereo failure

samples added

fix edgeboxes_demo

fix edgeboxes_demo

added edgeboxes bib

fix edgeboxes_demo

small fixes

fix edgeboxes_demo

fix warnings

fix warnings

small fixes

detectEdges needs rgb image instead bgr image.

Removed unnecessary protection

small fixes

added optimized anisotropic diffusion filter (rewrite of https://github.com/opencv/opencv/pull/8362)

fixed typo

replaces Mat& to Input/OutputArray in public methods.

small fix.

small fixes

1) Modified cv::ximgproc::niBlackThreshold to support binarization using popular Niblack inspired techniques. Sauvola, Wolf, and NICK support added.
2) Changed niblack_thresholding.cpp sample to allow experimentation with these new methods
3) Added/updated docs where appropriate

This simple detector runs up to two times faster than LSD while achieving
competitive results. Please evaluate the algorithm with samples/fld_lines.cpp

added thinning sample
added thinning tests

Paillou Filter

comment : english tab and assert  Mat(instead of new)

InputArray and ptr<Type>()

Outputarrays

OutputArray as argument

Update doc

- read image as 1-channel grayscale
- 3 trackers to control function parameters
- K allowed to go to negative values (usually to detect black text on white bg)

Now the filter natively supports StereoBM and StereoSGBM with no
parameter tuning required. Also, now user won't need to set the ROI and
the right matcher parameters manually, it is all done in the respective
convenience factory method based on the left matcher instance. Tutorial
was added to clarify the provided example of use.

All Niblack related code now resides inside ximgproc and not as a
separate new module.

DisparityWLSFilter demonstrated the best results, so I removed all the other
filters. Quality was significantly improved by adding confidence support
to the filter (left-right consistency + penalty for areas near depth discontinuities).
Filter was optimized using parallel_for_ and HAL intrinsics. Demo application was
rewritten for better compliance with OpenCV standards. Added accuracy and
performance tests. Documentation was added, as well as references to the
original papers.

modules/ximgproc/include/opencv2/ximgproc/fast_hough_transform.hpp moved from modules/ximgproc/include/opencv2/ximgproc.hpp
HoughPoint2Line returned Vec4i instead of OutputArray

Added basic interface and demo for disparity filtering, added unoptimized fast weighted least
squares filter implementation. Current demo tests domain transform, guided and
weighted least squares filters on a dataset, measures speed and quality.