Implementation of bit-exact resize. Internal calls to linear resize updated to use bit-exact version. (#9468)

If there are no OpenCL/UMat methods calls from application.

OpenCL subsystem is initialized:
- haveOpenCL() is called from application
- useOpenCL() is called from application
- access to OpenCL allocator: UMat is created (empty UMat is ignored) or UMat <-> Mat conversions are called

Don't call OpenCL functions if OPENCV_OPENCL_RUNTIME=disabled
(independent from OpenCL linkage type)

This statement was keeping HAAR cascades from leveraging opencl on nvidia devices. "localSize" on the featureEvaluator remains Size(0, 0) which sets the bool "use_ocl" to false. Adding this allows NVidia GPUs to leverage opencl HAAR Cascades

Updated integrations for:
cv::split
cv::merge
cv::insertChannel
cv::extractChannel
cv::Mat::convertTo - now with scaled conversions support
cv::LUT - disabled due to performance issues
Mat::copyTo
Mat::setTo
cv::flip
cv::copyMakeBorder - currently disabled
cv::polarToCart
cv::pow - ipp pow function was removed due to performance issues
cv::hal::magnitude32f/64f - disabled for <= SSE42, poor performance
cv::countNonZero
cv::minMaxIdx
cv::norm
cv::canny - new integration. Disabled for threaded;
cv::cornerHarris
cv::boxFilter
cv::bilateralFilter
cv::integral

clip the found objects in HOG as well (http://code.opencv.org/issues/3825); added test to check CascadeClassifier on small images (http://code.opencv.org/issues/3710)

improvements in Haar CascadeClassifier: 1) use CV_32S instead of CV_32F for the integral of squares (which is more accurate and more efficient); 2) skip the window if its contrast is too low

Avoid clearing HaarEvaluator::tofs variable (introduced fb8478eb) in HaarEvaluator::setImage(...) since it won't be recomputed by HaarEvaluator::computeOptFeatures() for cases where the input image dimensions and requested search scales are the same as the previous.  If none of the above change then it should be correct, and if they do change it will be computed correctly inside HaarEvaluator::computeOptFeatures() (in commit 952f9dbe)

I'm guessing there is a unit test for this that needs to be called 2x with the same image and search parameters (or another image with the same size) to illustrate the problem.

do not use this variable if it is not an tilted feature

Trying to fix whitespace issue.
This reverts commit dcc2512f.

added a check to prevent illegal access to operator[] when vector is empty

and unsynchronized.
Signed-off-by: Yan Wang <yan.wang@linux.intel.com>

layer_dy is not initialized correctly, it results in a small height
of the buffer, and so memory write is out of range in cv::integral.

Signed-off-by: Yan Wang <yan.wang@linux.intel.com>

- Gaussian blur;
- CascadeClassifier;
- Blenders in stitching module;
- Laplacial pyromids in stitching module.

the parameter variable.

It could improve the performance of
OCL_Cascade_Image_MinSize_CascadeClassifier.CascadeClassifier/*.
Especially,
OCL_Cascade_Image_MinSize_CascadeClassifier.CascadeClassifier/15
OCL_Cascade_Image_MinSize_CascadeClassifier.CascadeClassifier/16
could be improved about 2% in Intel platform.
Signed-off-by: Yan Wang <yan.wang@linux.intel.com>