- Use cv::AutoBuffer rather than new allocation
- Fix "assignment operator could not be generated" warning

mask = cv2.inRange(hsv, lower_green, upper_green)
to
mask = cv2.inRange(hsv, lower_blue, upper_blue)

It could improve performance when image size is large.
E.g. OCL_PyrUpFixture_PyrUp.PyrUp/18

This flag is already added in line 131.

In the unoptimized version of small symmetrical column filters, when we
try to detect if the kernel, ky, is equal to [1;2;1] or [1;-2;1] we
should take into consideration that the anchor points in the middle
element.

1. Remove unnecessary barriers.
2. Adjust CTA_SIZE based on the following cases for Intel platform:
   a) OCL_Photo_DenoisingGrayscale.DenoisingGrayscale
   b) OCL_Photo_DenoisingColored.DenoisingColored

Avoid stack overflow based on std::vector cause
std::vector::resize() segmentation fault.